Channel evaluation method for power line carrier communication and electronic equipment

Abstract

The invention provides a channel evaluation method for power line carrier communication and electronic equipment. The method comprises the following steps: acquiring a channel evaluation frame which is sent by a signal sending end through multiple spatial streams and comprises a time domain training signal and a time domain load signal; respectively executing fast Fourier transform on the two types of time domain signals to obtain corresponding frequency domain training signals and frequency domain load signals; estimating a first channel matrix of the frequency domain training signal, calculating a training signal residual error to identify an abnormal frequency domain symbol, processing to obtain an effective training signal, and estimating a second channel matrix and a noise covariance matrix of the effective training signal; constructing a channel equalization matrix based on the two matrixes, and respectively processing the two types of frequency domain signals to obtain a training signal signal to interference plus noise ratio and a load signal signal to noise ratio of each subcarrier of each spatial stream; and finally, carrying out weighted fusion on the two types of indexes to generate a fusion signal-to-noise ratio for evaluating the channel quality of each subcarrier of each spatial stream. According to the invention, robustness and accuracy of power line carrier communication channel evaluation can be considered.

Description

Technical Field

[0001] This application relates to the field of communication technology, and more specifically, to a channel evaluation method and electronic device for power line carrier communication. Background Technology

[0002] In communication systems, low-voltage power line broadband carrier communication relies on existing power distribution networks to achieve data transmission without the need for additional dedicated communication lines. It has significant advantages such as low cost and wide coverage, and has broad application prospects in fields such as smart grids and smart homes.

[0003] However, power lines were originally designed for power transmission, not data communication. The loads connected to these lines are diverse and time-varying, and the power grid generates strong electromagnetic noise interference during operation, resulting in an extremely harsh power line channel environment. Specifically, power line channels suffer from frequency-selective fading, with signals of different frequencies experiencing significantly different degrees of attenuation during transmission. They are also accompanied by complex interference forms such as narrowband interference and impulse / impact noise, and the channel characteristics dynamically change with load variations and power grid operating conditions, exhibiting strong time-varying characteristics. These channel defects cause severe distortion and bit errors during signal transmission, significantly reducing the reliability of data transmission.

[0004] Therefore, it is essential to conduct an accurate and robust assessment of channel quality before implementing power line carrier communication, so as to provide a key basis for selecting optimal transmission parameters and configuring spatial streams and subcarrier resources for the communication system. Summary of the Invention

[0005] Embodiments of this application provide a channel evaluation method, computer program product or computer program, computer-readable storage medium, or electronic device for power line carrier communication, which can at least to some extent balance the robustness and accuracy of power line carrier communication channel evaluation.

[0006] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part from practice of this application.

[0007] According to one aspect of the embodiments of this application, a channel evaluation method for power line carrier communication is provided. The method includes: acquiring a channel evaluation frame, the channel evaluation frame including a time-domain training signal and a time-domain payload signal transmitted by a signal transmitter through multiple spatial streams; performing a fast Fourier transform on the time-domain training signal and the time-domain payload signal to obtain a frequency-domain training signal and a frequency-domain payload signal, respectively, wherein the frequency-domain training signal corresponding to each spatial stream includes multiple repeated frequency-domain symbols, and different pseudo-random binary phase-shift keying modulation symbols are superimposed on the frequency-domain symbols of different spatial streams, and the frequency-domain payload signal corresponding to each spatial stream is filled with a known pseudo-random binary phase-shift keying sequence, wherein both the frequency-domain training signal and the frequency-domain payload signal are composed of multiple subcarriers; estimating the frequency-domain training signal... The system firstly constructs a first channel matrix and calculates a training signal residual based on the first channel matrix to identify anomalous frequency domain symbols in the frequency domain training signal according to the training signal residual. It then processes the anomalous frequency domain symbols in the frequency domain training signal to obtain an effective training signal and estimates a second channel matrix and a noise covariance matrix of the effective training signal. Based on the second channel matrix and the noise covariance matrix, it constructs a channel equalization matrix and processes the frequency domain training signal and the frequency domain payload signal based on the channel equalization matrix to obtain the training signal signal-to-interference-plus-noise ratio (SNR) and payload signal SNR for each subcarrier of each spatial stream. Finally, it performs weighted fusion of the training signal SNR and the payload signal SNR to obtain a fused SNR used to evaluate the channel quality of each subcarrier in each spatial stream.

[0008] In some embodiments of this application, based on the foregoing scheme, estimating the first channel matrix of the frequency domain training signal includes: based on locally stored known training signals, using a least squares estimation method, performing channel estimation and denoising on multiple frequency domain symbols in the frequency domain training signal to obtain the first channel matrix of the frequency domain training signal.

[0009] In some embodiments of this application, based on the foregoing scheme, the step of calculating the training signal residual based on the first channel matrix includes: determining training signal estimates corresponding to multiple frequency domain symbols based on the first channel matrix and the known training signal; calculating the deviation between each frequency domain symbol in the frequency domain training signal and the training signal estimate corresponding to each frequency domain symbol, and obtaining the training signal residual corresponding to each frequency domain symbol.

[0010] In some embodiments of this application, based on the foregoing scheme, the step of identifying anomalous frequency domain symbols in the frequency domain training signal according to the training signal residual includes: constructing an impulse noise detection statistic corresponding to each frequency domain symbol based on the training signal residual, and robustly estimating the robust centrality and robust scale of the impulse noise detection statistic using the median; comparing the impulse noise detection statistic with the robust centrality, the robust scale, and a preset threshold coefficient; if the absolute value of the difference between the impulse noise detection statistic and the robust centrality is greater than the product of the preset threshold coefficient and the robust scale, then determining that each frequency domain symbol is an anomalous frequency domain symbol.

[0011] In some embodiments of this application, based on the foregoing scheme, the processing of abnormal frequency domain symbols in the frequency domain training signal includes: removing the abnormal frequency domain symbols from the frequency domain training signal; or, assigning weights to each frequency domain symbol in the frequency domain training signal, wherein the weights assigned to the abnormal frequency domain symbols are lower than the weights assigned to the normal frequency domain symbols.

[0012] In some embodiments of this application, based on the foregoing scheme, estimating the second channel matrix and noise covariance matrix of the effective training signal includes: processing the known training signal according to the processing method for abnormal frequency domain symbols in the frequency domain training signal to obtain an effective known training signal; based on the effective known training signal, performing channel estimation and denoising on multiple frequency domain symbols in the effective training signal using a least squares estimation method to obtain the second channel matrix of the effective training signal; recalculating the training signal residual based on the effective training signal to obtain the noise covariance matrix; or, calculating the training signal residual in a weighted manner according to the weights of the effective training signal to obtain the noise covariance matrix.

[0013] In some embodiments of this application, based on the aforementioned scheme, the channel equalization matrix is ​​a minimum mean square error channel equalization matrix. The frequency domain training signal is processed based on the channel equalization matrix to obtain the signal-to-interference-plus-noise ratio (SIR) of the training signal for each subcarrier of each spatial stream. This includes: calculating an equalization correlation matrix based on the second channel matrix and the channel equalization matrix; calculating the expected power of each subcarrier of each spatial stream and the residual crosstalk power of each subcarrier of other spatial streams in the frequency domain training signal based on the equalization correlation matrix; calculating the noise power of each subcarrier of each spatial stream by combining the channel equalization matrix and the noise covariance matrix; and calculating the SIR of the training signal for each subcarrier of each spatial stream based on the expected power, the residual crosstalk power, and the noise power.

[0014] In some embodiments of this application, based on the aforementioned scheme, the frequency domain payload signal is processed based on the channel equalization matrix to obtain the signal-to-noise ratio (SNR) of the payload signal for each subcarrier in each spatial stream. This includes: equalizing the frequency domain payload signal based on the channel equalization matrix to obtain an equalized frequency domain payload signal; selecting subcarriers with a training signal-to-interference-plus-noise ratio (SNR) greater than a preset threshold as reliable subcarriers, and calculating a common error estimate based on the reliable subcarriers; performing common phase correction and common gain correction on the equalized frequency domain payload signal based on the common error estimate to obtain a corrected frequency domain payload signal; calculating the error vector amplitude of each subcarrier in each spatial stream based on the corrected frequency domain payload signal and the reference frequency domain payload signal, and converting the error vector amplitude into the SNR of the payload signal for each subcarrier in each spatial stream, wherein the time domain payload signal is obtained by the signal transmitting end performing a fast Fourier transform on the reference frequency domain payload signal.

[0015] In some embodiments of this application, based on the foregoing scheme, the weighted fusion of the training signal signal-to-interference-plus-noise ratio (SNR) and the payload signal SNR includes: determining a fusion weight based on the number of frequency domain symbols in the effective training signal or the proportion of abnormal frequency domain symbols, wherein the fusion weight is limited between a first set weight and a second set weight; and performing weighted fusion of the training signal SNR and the payload signal SNR according to the fusion weight.

[0016] According to one aspect of the embodiments of this application, a computer program product or computer program is provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods described in the above embodiments.

[0017] According to one aspect of the embodiments of this application, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the method described in the above embodiments.

[0018] According to one aspect of the embodiments of this application, an electronic device is provided, including: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to perform the method described in the above embodiments.

[0019] Based on the technical solution proposed in this application, by introducing multiple spatial streams to transmit training and payload signals in parallel, and combining this with the differentiated design of pseudo-random binary phase-shift keying modulation symbols, the efficiency of signal transmission and anti-interference capability can be improved. By identifying and processing anomalous frequency domain symbols, the impact of impulse noise, strong interference, and other factors on channel evaluation can be reduced, making the estimation of the second channel matrix and noise covariance matrix more accurate. By equalizing the signal using the optimized channel equalization matrix, channel distortion can be effectively compensated, and signal quality improved. By weighted fusion of the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal, the limitations of a single evaluation method can be avoided. The final fused SNR can comprehensively, accurately, and robustly reflect the channel quality of each spatial stream and each subcarrier, providing a reliable basis for the communication system to select optimal transmission parameters and configure spatial stream and subcarrier resources.

[0020] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0021] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort. In the drawings: Figure 1 A flowchart of a channel evaluation method for power line carrier communication according to an embodiment of this application is shown; Figure 2 A detailed flowchart of a channel evaluation method for power line carrier communication according to an embodiment of this application is shown; Figure 3 A schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application is shown. Detailed Implementation

[0022] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art.

[0023] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0024] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0025] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined. Therefore, the actual execution order may change depending on the actual situation.

[0026] It should be noted that "multiple" in this article refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0027] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such uses of these terms can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described.

[0028] In communication systems, low-voltage power line broadband carrier communication relies on existing power distribution networks to achieve data transmission without the need for additional dedicated communication lines. It has significant advantages such as low cost and wide coverage, and has broad application prospects in fields such as smart grids and smart homes.

[0029] However, power lines were originally designed for power transmission, not data communication. The loads connected to these lines are diverse and time-varying, and the power grid generates strong electromagnetic noise interference during operation, resulting in an extremely harsh power line channel environment. Specifically, power line channels suffer from frequency-selective fading, with signals of different frequencies experiencing significantly different degrees of attenuation during transmission. They are also accompanied by complex interference forms such as narrowband interference and impulse / impact noise, and the channel characteristics dynamically change with load variations and power grid operating conditions, exhibiting strong time-varying characteristics. These channel defects cause severe distortion and bit errors during signal transmission, significantly reducing the reliability of data transmission. Therefore, before implementing power line carrier communication, it is essential to conduct a precise and robust assessment of channel quality to provide a crucial basis for selecting optimal transmission parameters and configuring spatial streams and subcarrier resources for the communication system.

[0030] The inventors of this application have discovered that in the prior art, channel assessment methods based on training field (TF) signals are susceptible to impulse noise, leading to instantaneous distortion of the training signal, causing noise estimation bias, and consequently introducing systematic errors in the signal-to-interference-plus-noise ratio (SNR). While SNR assessment based on payload (PL) signals is closer to demodulation capability, common phase error, common gain error, and impulse noise contamination may cause the error vector amplitude to be systematically amplified or exhibit abnormal fluctuations, resulting in insufficient accuracy of the channel assessment results and making it difficult to meet the high-precision requirements of communication systems for channel quality assessment.

[0031] In this context, this application proposes a channel evaluation scheme for power line carrier communication to balance the robustness and accuracy of power line carrier communication channel evaluation.

[0032] The implementation details of the technical solutions in the embodiments of this application are described below: See Figure 1 The diagram shows a flowchart of a channel evaluation method for power line carrier communication according to an embodiment of the present application, the method being executed by a computing device with computing capabilities.

[0033] like Figure 1 As shown, the channel evaluation method for power line carrier communication includes at least steps 110 to 160, which are described in detail below: In step 110, a channel evaluation frame is obtained, which includes a time-domain training signal and a time-domain payload signal transmitted by the signal transmitter through multiple spatial streams.

[0034] In this application, the channel evaluation frame is a specific frame structure generated by the signal transmitter specifically for channel evaluation. It includes a time-domain training signal and a time-domain payload signal, and is transmitted through multiple spatial streams. Here, a spatial stream refers to an independent data transmission path from the signal transmitter to the signal receiver in a Multiple-Input Multiple-Output (MIMO) communication system. Multiple spatial streams enable parallel data transmission, improving the communication rate. The time-domain training signal provides basic data for channel matrix estimation, while the time-domain payload signal is used to assist in verifying channel quality; the two complement each other.

[0035] Continue to refer to Figure 1 In step 120, the time-domain training signal and the time-domain payload signal are subjected to fast Fourier transform to obtain the frequency-domain training signal and the frequency-domain payload signal, respectively. The frequency-domain training signal corresponding to each spatial stream includes multiple repeated frequency-domain symbols, and different pseudo-random binary phase-shift keying modulation symbols are superimposed on the frequency-domain symbols of different spatial streams. The frequency-domain payload signal corresponding to each spatial stream is filled with a known pseudo-random binary phase-shift keying sequence. Both the frequency-domain training signal and the frequency-domain payload signal are composed of multiple subcarriers.

[0036] In this application, the Fast Fourier Transform (FFT) is a key processing method for converting time-domain signals into frequency-domain signals. Through this transformation, signals that are difficult to analyze in the time domain can be decomposed into frequency-domain signals with different frequency components, which facilitates subsequent evaluation of the channel characteristics of different subcarriers.

[0037] In this application, the frequency domain training signal corresponding to each spatial stream contains multiple repeated frequency domain symbols. The repeated design can improve the reliability of channel estimation and reduce the impact of random noise.

[0038] In this application, different pseudo-random binary phase shift keying (BPSK) modulation symbols are superimposed on the frequency domain symbols of different spatial streams, which can distinguish the signals of different spatial streams and avoid interference between spatial streams. The pseudo-random BPSK modulation symbols have good autocorrelation and cross-correlation, which can effectively improve the accuracy of signal recognition.

[0039] In this application, the frequency domain payload signal is filled with a known pseudo-random BPSK sequence, which enables the receiver to compare and analyze the received signal with the known sequence, thereby achieving indirect assessment of channel quality. The design of the known sequence ensures the feasibility and accuracy of the comparative analysis.

[0040] In this application, both the frequency domain training signal and the frequency domain payload signal are composed of multiple subcarriers. Subcarriers are the basic transmission units in Orthogonal Frequency Division Multiplexing (OFDM) technology. Different subcarriers are orthogonal in frequency and can transmit data in parallel. The channel characteristics of each subcarrier are different, so it is necessary to evaluate the channel quality of each subcarrier separately.

[0041] Next, this application will describe steps 110 to 110 above in conjunction with a specific embodiment: For example, in a specific embodiment, the signal transmitting end generates and transmits a channel evaluation frame, and the signal receiving end performs FFT processing on the time-domain training signal and time-domain payload signal of the received channel evaluation frame to obtain the frequency-domain training signal and frequency-domain payload signal.

[0042] In this embodiment, it is assumed that: This represents the set of transmitting ports (i.e., transmitting antennas) in a signal transmitter. Indicates the first One transmission port; This refers to the set of receiving ports (i.e., receiving antennas) at the signal receiving end. Indicates the first One receiving port; A set representing spatial flows, Indicates the first The first spatial flow; This represents the set of frequency domain symbols in the frequency domain training signal. In the frequency domain training signal, the first... One frequency domain symbol; This represents the set of frequency domain symbols in the frequency domain payload signal. In the frequency domain load signal, the first One frequency domain symbol; Represents the set of subcarriers. Indicates the first Subcarriers.

[0043] In this embodiment, the frequency domain training signal consists of repeating frequency domain symbols, which are generated based on a known phase table of the training signal. Furthermore, each frequency domain symbol in the training signal is superimposed with a different BPSK modulation symbol according to a different spatial flow. Therefore, the subcarrier... The corresponding reference frequency domain training signal is:

[0044] The time-domain training signal of the channel evaluation frame is obtained by the signal transmitting end performing a fast Fourier transform on the reference frequency-domain training signal. It can be represented as:

[0045] in, For frequency domain symbol phase, For subcarriers In space flow Frequency domain symbols BPSK modulation symbols superimposed on top.

[0046] In this embodiment, the frequency domain symbols of the frequency domain payload signal are filled with a known BPSK sequence. Therefore, the subcarrier... The corresponding reference frequency domain load signal is:

[0047] The time-domain payload signal of the channel evaluation frame is obtained by the signal transmitting end performing a fast Fourier transform on the reference frequency-domain payload signal. It can be represented as:

[0048] in, For subcarriers In space flow Frequency domain symbols A known pseudo-random BPSK sequence filled with top padding.

[0049] Continue to refer to Figure 1 In step 130, the first channel matrix of the frequency domain training signal is estimated, and the training signal residual is calculated based on the first channel matrix to identify abnormal frequency domain symbols in the frequency domain training signal according to the training signal residual.

[0050] In this application, the estimation of the first channel matrix is ​​a preliminary process for obtaining channel transmission characteristics. The channel matrix can reflect the transmission gain and phase offset of the signal among various transmit antennas, receive antennas, spatial streams, and subcarriers. The training signal residual is calculated based on the first channel matrix. The residual is the difference between the actually received frequency domain training signal and the training signal estimated based on the first channel matrix. This difference is mainly caused by factors such as noise and interference. By analyzing the residual, severely interfered anomalous frequency domain symbols can be identified. Anomalous frequency domain symbols typically refer to frequency domain symbols that have suffered severe signal distortion due to impulse noise, strong interference, etc.

[0051] In such Figure 1In step 130 shown, estimating the first channel matrix of the frequency domain training signal can be performed according to step 131 as follows: Step 131: Based on the known training signal stored locally, the least squares estimation method is used to perform channel estimation and denoising on multiple frequency domain symbols in the frequency domain training signal to obtain the first channel matrix of the frequency domain training signal.

[0052] In this application, the locally stored known training signal is consistent with the original signal (the signal that has not been transmitted through the channel and is not subject to noise interference) of the frequency domain training signal sent by the signal transmitter. The receiver stores the known training signal in advance to provide a reference benchmark for channel estimation, so that the receiver can deduce the transmission characteristics of the channel by comparing the known training signal with the actual received frequency domain training signal.

[0053] In this application, least squares estimation is a commonly used parameter estimation method. Its core idea is to find the parameter estimate that minimizes the sum of squared errors between the actual observations and the predicted values ​​based on the estimated parameters. In channel estimation, least squares estimation determines the estimated value of the channel matrix by minimizing the sum of squared errors between the actually received frequency domain training signal and the predicted signal based on the estimated channel matrix and the known training signal.

[0054] In this application, the channel estimation results for a single frequency domain symbol are significantly affected by noise, resulting in low accuracy. By performing channel estimation on multiple frequency domain symbols separately, multiple channel matrix estimates can be obtained. Then, by employing a merging process (such as averaging or weighted averaging), the random influence of noise can be reduced, improving the accuracy and robustness of the channel matrix estimation. In other words, the essence of merging and denoising is to utilize the statistical characteristics of multiple observations to offset the random fluctuations of noise, making the estimation results closer to the true channel characteristics.

[0055] Next, this application will describe step 131 based on the above embodiments: In this embodiment, the first channel matrix satisfies the least squares form. A more accurate channel estimate is obtained by denoising and combining channel estimates of the same subcarriers in multiple frequency domain symbols. Assuming the subcarriers... The frequency domain training signal is:

[0056] subcarrier First channel matrix It can be represented as:

[0057] in, Represents the complex conjugate transpose of a matrix; This represents the inverse of a matrix.

[0058] Based on step 131 above, the feasibility and accuracy of channel estimation can be ensured. The consistency between the known training signal and the original training signal at the transmitting end provides a reliable basis for the receiving end to infer channel characteristics. Choosing the least squares estimation method balances estimation accuracy and computational complexity, enabling efficient implementation in practical power line carrier communication applications and meeting the system's requirements for real-time performance and computational resources. Channel estimation and noise reduction of multiple frequency domain symbols effectively reduces the impact of random noise on the channel estimation results. While the channel estimation result of a single frequency domain symbol is significantly affected by noise fluctuations, the merging of multiple frequency domain symbols can utilize the randomness of noise to partially cancel it out. This makes the estimation result of the first channel matrix more stable and closer to the true channel characteristics, providing high-quality basic data for subsequent steps such as training signal residual calculation and abnormal frequency domain symbol identification, thereby improving the accuracy and robustness of the entire channel evaluation method.

[0059] In such Figure 1 In step 130 shown, the calculation of the training signal residual based on the first channel matrix can be performed according to steps 132 to 133 as follows: Step 132: Based on the first channel matrix and the known training signal, determine the training signal estimates corresponding to multiple frequency domain symbols.

[0060] Step 133: Calculate the deviation between each frequency domain symbol in the frequency domain training signal and the training signal estimate corresponding to each frequency domain symbol, and obtain the training signal residual corresponding to each frequency domain symbol.

[0061] In this application, the calculation of the training signal estimation follows a signal transmission model that considers the influence of the channel matrix on the signal gain and phase shift, accurately reflecting the signal transmission pattern in the channel. The training signal residual is obtained by comparing the actual received frequency-domain training signal with the estimated training signal and calculating the deviation between the two.

[0062] In this application, the magnitude of the training signal residual directly reflects the degree of difference between the actual received signal and the ideal received signal. The greater the difference, the more severe the impact of noise, interference, etc., on the signal during transmission. Each frequency domain symbol corresponds to an independent training signal residual, which allows the receiver to analyze the interference situation of each frequency domain symbol separately, providing an accurate basis for subsequent identification of abnormal frequency domain symbols. The residuals of different frequency domain symbols may vary significantly because noise and interference in the power line channel are time-varying and random; the intensity and type of interference may differ at different times (corresponding to different frequency domain symbols).

[0063] Next, this application will describe steps 132 to 133 based on the above embodiments: In this embodiment, the training signal residual corresponding to each frequency domain symbol is calculated based on the first channel matrix. Assuming subcarriers... The corresponding training signal residual is Then frequency domain symbols Corresponding training signal residual It can be represented as:

[0064] Based on steps 132 and 133 above, by determining the training signal estimate based on the first channel matrix and the known training signal, the transmission process of the signal under ideal channel conditions can be accurately simulated, providing a reliable comparison benchmark for residual calculation. The accuracy of the training signal estimate directly affects the reliability of the residual calculation, and the explicit calculation logic in this step ensures the rationality of the training signal estimate. By calculating the deviation between each frequency domain symbol and the corresponding training signal estimate, the training signal residual can be obtained, enabling precise quantification of the interference situation of each frequency domain symbol. The magnitude of the residual can intuitively reflect the degree of signal distortion, providing a scientific and accurate basis for subsequent identification of abnormal frequency domain symbols. Calculating the residual independently for each frequency domain symbol allows the receiver to analyze the channel transmission situation of different frequency domain symbols in detail, avoiding the loss of details caused by overall analysis, thereby improving the accuracy of abnormal frequency domain symbol identification. This lays a solid foundation for the subsequent acquisition of effective training signals and the accurate estimation of the second channel matrix and noise covariance matrix, further improving the robustness and accuracy of the entire channel evaluation method.

[0065] In such Figure 1 In step 130 shown, the step of identifying abnormal frequency domain symbols in the frequency domain training signal based on the training signal residual can be performed according to the following steps 134 to 136: Step 134: Construct the impulse noise detection statistic corresponding to each frequency domain symbol based on the training signal residual, and robustly estimate the robust centrality and robust scale of the impulse noise detection statistic using the median.

[0066] Step 135: Compare the impact noise detection statistics with the robust center quantity, the robust scale quantity, and the preset threshold coefficient.

[0067] Step 136: If the absolute value of the difference between the impact noise detection statistic and the robustness center quantity is greater than the product of the preset threshold coefficient and the robustness scale quantity, then each frequency domain symbol is determined to be an abnormal frequency domain symbol.

[0068] In this application, the impulse noise detection statistic typically employs the cross-subcarrier aggregated energy of the training signal residuals. This is calculated by summing the squares (or the square of the Euclidean norm) of the residuals of all subcarriers for each frequency domain symbol. A larger aggregated energy indicates a more severe impact of noise and interference on the frequency domain symbol. This construction method comprehensively integrates the residual information of all subcarriers for each frequency domain symbol, accurately reflecting the overall interference situation of that frequency domain symbol.

[0069] In this application, the robust centrality and robust scaling measures of the impulse noise detection statistic are robustly estimated using the median. This is because traditional mean and variance estimations are easily affected by outliers (i.e., the detection statistics corresponding to outlier frequency domain symbols), leading to significant estimation biases. The median, on the other hand, exhibits strong robustness and is less susceptible to outlier interference. The robust centrality measures the central tendency of the impulse noise detection statistic, reflecting the average level of the normal frequency domain symbol detection statistic. The robust scaling measures the dispersion of the impulse noise detection statistic, reflecting the fluctuation range of the normal frequency domain symbol detection statistic.

[0070] In this application, the impulse noise detection statistics of all frequency domain symbols can be sorted by size, and the value at the middle position is taken as the robust central quantity (i.e., the median). The absolute value of the difference between each detection statistic and the robust central quantity is calculated, and the median of these absolute values ​​is taken. This median is then multiplied by a preset coefficient (used to convert the median deviation into an approximate standard deviation) to obtain the robust scale quantity. This estimation method can effectively eliminate the influence of anomalous frequency domain symbols on the estimation of the central and scale quantities, ensuring that the estimation results can accurately reflect the statistical characteristics of normal frequency domain symbols.

[0071] In this application, the impulse noise detection statistics for each frequency domain symbol are compared with robust center quantity, robust scale quantity, and preset threshold coefficient to establish reasonable judgment rules, enabling accurate differentiation between normal and abnormal frequency domain symbols. The preset threshold coefficient is an empirical value determined based on actual application scenarios and channel characteristics. It is used to adjust the sensitivity of anomaly judgment. A larger threshold coefficient results in stricter anomaly judgment and a lower probability of misjudging anomalies, but may miss some slightly anomalous frequency domain symbols. A smaller threshold coefficient results in more sensitive anomaly judgment and a lower probability of missed judgment, but may misjudge some normal frequency domain symbols. The optimal preset threshold coefficient can be determined through experiments and real-world scenario testing to achieve a balance between the misjudgment rate and the missed judgment rate.

[0072] In this application, when the absolute value of the difference between the impulse noise detection statistic and the robustness center quantity is greater than the product of the preset threshold coefficient and the robustness scale quantity, it indicates that the interference level of the frequency domain symbol significantly exceeds the fluctuation range of the normal frequency domain symbol, and therefore the frequency domain symbol is determined to be an abnormal frequency domain symbol. This judgment rule is based on statistical theory, has strong scientific validity and reliability, and can effectively identify abnormal frequency domain symbols affected by impulse noise and strong interference.

[0073] Next, this application will describe steps 134 to 136 based on the above embodiments: In this embodiment, to prevent the influence of impulse interference, the impulse noise detection statistic is the cross-subcarrier aggregated energy. The accuracy of noise variance estimation is improved by robustly estimating the noise variance using the median. Frequency domain symbol. Impact noise detection statistics It can be represented as:

[0074] in, This represents the Euclidean norm of the matrix. The set of anomalous frequency domain symbols identified based on impulse noise detection statistics includes the following steps: The first step is to use frequency domain symbols. Impact noise detection statistics Calculate and estimate robust central quantities and estimating robust scalar quantities , and It can be represented as:

[0075] in, This indicates the calculation of the median.

[0076] The second step, according to , and Make a decision for each frequency domain symbol: For each frequency domain symbol If satisfied Then the frequency domain symbol is considered to be It is an abnormal frequency domain symbol, that is .in, This is the threshold coefficient.

[0077] Based on steps 134 to 136 above, by constructing impulse noise detection statistics based on training signal residuals, a comprehensive quantification of the interference level of each frequency domain symbol can be achieved, avoiding the limitations of single subcarrier residual judgment and comprehensively reflecting the overall interference situation of frequency domain symbols. Using median robust estimation of robust centrality and robust scale effectively overcomes the shortcomings of traditional mean and variance estimation, which are easily affected by outliers, ensuring that the centrality and scale accurately reflect the statistical characteristics of normal frequency domain symbols and providing a reliable reference benchmark for anomaly judgment. By setting reasonable judgment rules, comparing the impulse noise detection statistics with robust centrality, robust scale, and preset threshold coefficients, accurate identification of abnormal frequency domain symbols can be achieved. This judgment rule is based on statistical theory, has strong scientific validity and adaptability, and can adjust the preset threshold coefficients according to actual scenarios, achieving a balance between false positive and false negative rates. Overall, the abnormal frequency domain symbol identification logic has high accuracy and robustness, and can effectively identify abnormal frequency domain symbols affected by impulse noise and strong interference. This lays a key foundation for subsequent abnormal symbol processing and improving the accuracy of channel assessment, thereby enhancing the adaptability of the entire channel assessment method to complex and harsh power line channel environments.

[0078] Continue to refer to Figure 1 In step 140, the abnormal frequency domain symbols in the frequency domain training signal are processed to obtain the effective training signal, and the second channel matrix and noise covariance matrix of the effective training signal are estimated.

[0079] In this application, an effective training signal is obtained after processing the anomalous frequency domain symbols. The effective training signal can more realistically reflect the actual transmission characteristics of the channel. The second channel matrix and noise covariance matrix re-estimated based on the effective training signal have higher accuracy and robustness compared to the first channel matrix and the initial noise covariance matrix (estimated values ​​without processing anomalous symbols). The noise covariance matrix can reflect the statistical characteristics of noise across different receiving antennas and spatial flows, and is an important parameter for subsequent construction of the channel equalization matrix and calculation of the signal-to-noise ratio.

[0080] In such Figure 1 In step 140, the processing of abnormal frequency domain symbols in the frequency domain training signal can be performed according to either step 141 or step 142: Step 141: Remove the abnormal frequency domain symbols from the frequency domain training signal.

[0081] In this application, identified aberrant frequency domain symbols can be directly removed from the frequency domain training signal set. Subsequent processes such as channel matrix reestimation and noise covariance matrix estimation only use the remaining normal frequency domain symbols (i.e., valid training signals). The core idea of ​​this approach is to completely eliminate the interference of aberrant frequency domain symbols. Because aberrant frequency domain symbols are severely affected by noise and interference, the channel information they contain is seriously distorted, and continued use will lead to significant deviations in subsequent estimation results. After removing aberrant symbols, the remaining normal frequency domain symbols can more realistically reflect the actual transmission characteristics of the channel, and estimations and calculations based on these symbols have higher reliability. For example, if the frequency domain training signal contains 8 frequency domain symbols and 1 aberrant frequency domain symbol is identified, then the remaining 7 normal frequency domain symbols are used as valid training signals, and subsequent processes are based solely on these 7 frequency domain symbols.

[0082] Based on step 141 above, by removing abnormal frequency domain symbols, the interference of abnormal symbols on the subsequent evaluation process can be completely eliminated, ensuring the purity of the effective training signal. This makes the estimation of the channel matrix and noise covariance matrix based on the effective training signal more accurate, especially suitable for scenarios where abnormal symbols are severely distorted and have a significant impact on the evaluation results. This method is simple, direct, and easy to implement, and can quickly improve the robustness of the evaluation results.

[0083] Step 142: Assign weights to each frequency domain symbol in the frequency domain training signal, wherein the weights assigned to the abnormal frequency domain symbols are lower than the weights assigned to the normal frequency domain symbols.

[0084] In this application, weights can also be assigned to each frequency domain symbol, with lower weights assigned to anomalous symbols and higher weights to normal symbols. The advantage of this approach is that it fully utilizes the information from all frequency domain symbols, avoiding data waste caused by removing anomalous symbols. This is particularly suitable for scenarios with a large number of anomalous symbols or mild anomalousness. The weights are set based on the degree of anomalousness of the frequency domain symbol; the more severe the anomalousness, the lower the weight, thus reducing its impact on the overall estimation result. Normal symbols have higher weights, fully leveraging their accurate reflection of channel characteristics. The specific values ​​of the weights can be determined in various ways, such as based on the deviation between the impulse noise detection statistic and the robustness center quantity; the larger the deviation, the smaller the weight. Alternatively, a fixed weight allocation rule can be used, such as a weight of 1 for normal symbols and 0.1 for anomalous symbols. In subsequent processes such as channel matrix reestimation and noise covariance matrix estimation, a weighted calculation method is used, multiplying the signal value of each frequency domain symbol by its corresponding weight before merging the results, thereby suppressing the influence of anomalous symbols.

[0085] Based on step 142 above, by setting weights for frequency domain symbols, the adverse effects of anomalous symbols can be suppressed, and the information of all symbols can be fully utilized to avoid data waste. This method is suitable for scenarios with a large number of anomalous symbols or a mild degree of anomalousness. Through reasonable weight allocation, interference can be reduced while retaining effective information, thereby improving the accuracy of evaluation results and data utilization.

[0086] Both processing methods described above are designed for the characteristics of abnormal frequency domain symbols. They can be flexibly selected according to actual application scenarios, effectively solving the evaluation bias problem caused by abnormal frequency domain symbols. This provides high-quality and effective training signals for the accurate estimation of the second channel matrix and noise covariance matrix, thereby promoting the performance improvement of the entire channel evaluation method.

[0087] In such Figure 1 In step 140, estimating the second channel matrix and noise covariance matrix of the effective training signal can be performed according to steps 143 to 145 as follows: Step 143: Process the known training signal according to the processing method for abnormal frequency domain symbols in the frequency domain training signal to obtain a valid known training signal.

[0088] Step 144: Based on the effective known training signal, the least squares estimation method is used to perform channel estimation and denoising on multiple frequency domain symbols in the effective training signal to obtain the second channel matrix of the effective training signal.

[0089] Step 145: Recalculate the training signal residual based on the effective training signal to obtain the noise covariance matrix; or, calculate the training signal residual in a weighted manner according to the weights of the effective training signal to obtain the noise covariance matrix.

[0090] In this application, since the effective training signal is the frequency domain training signal after processing anomalous symbols, its corresponding known training signal also needs to be processed in the same way to ensure the matching between the two, thereby ensuring the accuracy of channel estimation. If the anomalous symbol removal processing method is adopted, then the portion corresponding to the anomalous symbol in the known training signal also needs to be removed; that is, the effective known training signal is the set of frequency domain symbols in the known training signal that correspond to the effective training signal. If the weighting processing method is adopted, then the same weight needs to be assigned to the corresponding frequency domain symbols in the known training signal; the effective known training signal is the set of known training signals with weight labels. For example, if the third frequency domain symbol is removed from the frequency domain training signal, then the third frequency domain symbol is also removed from the known training signal to obtain the effective known training signal. If the frequency domain training signal assigns a weight to each symbol, then each symbol in the known training signal is also assigned the same weight.

[0091] In this application, based on the effective known training signal, a second channel matrix is ​​obtained by performing channel estimation and denoising on multiple frequency domain symbols in the effective training signal using least squares estimation. This process is similar to the estimation process of the first channel matrix, but the input data becomes the effective training signal and the effective known training signal. Since the effective training signal has eliminated or suppressed the interference of anomalous symbols, the channel estimation result based on it is more accurate. Specifically, for the case where anomalous symbols are removed, both the effective training signal and the effective known training signal contain only normal frequency domain symbols. Least squares estimation is used to perform channel estimation on each normal frequency domain symbol, followed by denoising and merging to obtain the second channel matrix. For the case where weights are set, a weighted least squares estimation method is used. The channel estimate value of each frequency domain symbol is multiplied by its corresponding weight before denoising and merging to obtain the second channel matrix. The weighted processing can further highlight the role of normal symbols and suppress the residual influence of anomalous symbols.

[0092] In this application, the noise covariance matrix is ​​estimated based on the training signal residuals recalculated from the effective training signals, or on the training signal residuals calculated using weighted averages of the effective training signals. The recalculated training signal residuals are obtained based on the second channel matrix and the effective known training signals; that is, the difference between the effective training signals and the training signals estimated based on the second channel matrix and the effective known training signals is mainly caused by noise, thus accurately reflecting the statistical characteristics of the noise. For cases where outlier symbols are removed, the residuals are directly calculated based on the effective training signals and the effective known training signals, and then the noise covariance matrix is ​​estimated based on these residuals. The noise covariance matrix can be estimated using the outer product average of the residuals. For cases where weights are set, a weighted outer product average can be used to estimate the noise covariance matrix. This weighted estimation method can adjust the contribution of the residuals to the noise covariance matrix estimation according to the symbol weights, further improving the accuracy of the estimation.

[0093] Next, this application will describe step 143 or step 145 based on the above embodiments: In this embodiment, it is assumed that the set of frequency domain symbols in the effective training signal is as follows: The number of symbols in the frequency domain is , subcarrier The corresponding effective training signal is:

[0094] The effective known training signals are:

[0095] The second channel matrix of the effective training signal It can be represented as:

[0096] In one scenario, the noise covariance matrix can be obtained by recalculating the training signal residual based on the effective training signal. Here, the subcarrier... The corresponding noise covariance matrix It can be represented as:

[0097] In one scenario, the noise covariance matrix can be calculated using a weighted average based on the frequency domain symbol weights. Assume the frequency domain symbols... The weight is ,but It can be represented as:

[0098] Based on steps 143 to 145 above, by processing the known training signal in the same way as the frequency domain training signal, the matching between the effective training signal and the effective known training signal can be ensured, providing reliable input data for the accurate estimation of the second channel matrix. Employing least squares estimation combined with merging and denoising processing can fully utilize the information from multiple frequency domain symbols in the effective training signal, reducing the impact of noise on channel estimation. The resulting second channel matrix has higher accuracy and robustness compared to the first channel matrix, and more realistically reflects the actual transmission characteristics of the channel. The estimation of the noise covariance matrix is ​​based on the residuals or weighted residuals of the effective training signal, matching the estimation of the second channel matrix. This ensures that the noise covariance matrix accurately reflects the noise statistical characteristics of the current channel, providing key parameters for the subsequent construction of the channel equalization matrix and the calculation of the signal-to-noise ratio. Overall, the above steps guarantee the consistency and accuracy of the estimation of the second channel matrix and the noise covariance matrix, effectively overcoming the adverse effects of abnormal frequency domain symbols, and providing a solid guarantee for the smooth progress of the subsequent channel evaluation process and the accuracy of the evaluation results.

[0099] Continue to refer to Figure 1 In step 150, a channel equalization matrix is ​​constructed based on the second channel matrix and the noise covariance matrix, and the frequency domain training signal and the frequency domain payload signal are processed based on the channel equalization matrix to obtain the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal for each subcarrier of each spatial stream.

[0100] In this application, the channel equalization matrix compensates for channel distortion and reduces the impact of noise and interference on signal transmission. The channel equalization matrix constructed based on the second channel matrix and the noise covariance matrix can achieve optimal equalization processing for the current channel characteristics. By processing the frequency domain training signal and the frequency domain payload signal using this channel equalization matrix, the signal-to-interference-plus-noise ratio (SINR) of the training signal and the signal-to-noise ratio (SNR) of the payload signal, which reflect channel quality, can be obtained. SINR is the ratio of signal power to interference plus noise power, while SNR is the ratio of signal power to noise power. Both are important indicators for measuring channel quality. SINR can more comprehensively consider interference factors and is suitable for scenarios with severe interference, while SNR focuses more on reflecting the impact of noise on the signal.

[0101] In such Figure 1 In step 150, the channel equalization matrix is ​​a minimum mean square error channel equalization matrix. Based on the channel equalization matrix, the frequency domain training signal is processed to obtain the signal-to-interference-plus-noise ratio (SNR) of the training signal for each subcarrier in each spatial stream. This can be performed according to steps 151 to 154 as follows: Step 151: Calculate the equalization correlation matrix based on the second channel matrix and the channel equalization matrix.

[0102] Step 152: Calculate the expected power of each subcarrier of each spatial stream and the residual crosstalk power of each subcarrier of other spatial streams in the frequency domain training signal based on the equalization correlation matrix.

[0103] Step 153: Calculate the noise power of each subcarrier in each spatial stream by combining the channel equalization matrix and the noise covariance matrix.

[0104] Step 154: Calculate the training signal-to-interference-plus-noise ratio (SIR) for each subcarrier of each spatial stream based on the expected power, the residual crosstalk power, and the noise power.

[0105] In this application, the channel equalization matrix adopts the minimum mean square error (MMSE) channel equalization matrix. The goal of MMSE equalization is to minimize the mean square error between the equalized signal and the ideal signal. Compared with other equalization methods, MMSE equalization can better balance channel distortion compensation and noise suppression, and is especially suitable for power line channel environments with severe noise and interference.

[0106] In this application, the equalization correlation matrix reflects the correlation between the equalized signal and the various spatial flows, and is the basis for calculating the expected power and residual crosstalk power.

[0107] In this application, the desired power is the power of the spatial flow signal itself after equalization, reflecting the signal strength. Residual crosstalk power is the interference power of other spatial flows on this spatial flow. The calculation of desired power and residual crosstalk power can accurately quantify the signal strength itself and the degree of interference between spatial flows.

[0108] In this application, noise power reflects the intensity of the equalized noise.

[0109] Next, this application will describe step 151 or step 154 ​​based on the above embodiments: In this embodiment, to suppress noise amplification at deep fading subcarriers during channel equalization, channel effects are eliminated through minimum mean square error channel equalization, based on the second channel matrix. and noise covariance matrix Construct the minimum mean square error channel equalization matrix. Then the subcarriers... Minimum mean square error channel equalization matrix It can be represented as:

[0110] in, It is an identity matrix.

[0111] According to the second channel matrix Noise covariance matrix and minimum mean square error channel equalization matrix Calculating the signal-to-interference-plus-noise ratio (SIR) of the training signal for each subcarrier in each spatial stream of the frequency domain training signal includes the following steps: The first step is for subcarriers Using the second channel matrix and noise covariance matrix The equilibrium correlation matrix is ​​calculated. , can be represented as:

[0112] The second step is to use the equilibrium correlation matrix. The spatial flow in the frequency domain training signal can be calculated. subcarrier Expected power And in addition to spatial flow Frequency domain training signals of other spatial flows besides those in the spatial flow Residual crosstalk power , and They can be represented as:

[0113] The third step is to use the minimum mean square error channel equalization matrix. and noise covariance matrix The spatial flow in the frequency domain training signal is calculated. subcarrier noise power , can be represented as:

[0114] in, for The OK.

[0115] Fourth step, according to , and The signal-to-interference-plus-noise ratio (SIR) of the training signal for each subcarrier in each spatial stream of the frequency domain training signal can be calculated. , can be represented as:

[0116] Based on steps 151 to 154 above, selecting the minimum mean square error channel equalization matrix can effectively suppress noise amplification while compensating for channel distortion, adapting to the severe noise and interference characteristics of power line channels, and providing a good signal foundation for signal-to-interference-plus-noise ratio (SIR) calculation. By calculating the equalization correlation matrix, the gain of the equalized signal and the crosstalk gain between spatial flows can be accurately quantified, providing a reliable basis for calculating the expected power and residual crosstalk power. The calculation process for expected power, residual crosstalk power, and noise power is scientifically sound and can accurately reflect signal strength, spatial flow interference intensity, and noise intensity, respectively, comprehensively considering key factors affecting channel quality. The SIR of the training signal calculated based on these three parameters can comprehensively and accurately measure the channel quality of each spatial flow and each subcarrier, providing a high-quality evaluation index for subsequent channel assessment and fusion, effectively addressing the complex characteristics of power line channels, and improving the accuracy and reliability of channel assessment.

[0117] In such Figure 1 In step 150, the frequency domain payload signal is processed based on the channel equalization matrix to obtain the signal-to-noise ratio of the payload signal for each subcarrier in each spatial stream. This can be performed according to steps 155 to 158 as follows: Step 155: Perform equalization processing on the frequency domain load signal based on the channel equalization matrix to obtain an equalized frequency domain load signal.

[0118] Step 156: Select subcarriers with a signal-to-interference-plus-noise ratio (SNR) greater than a preset threshold as reliable subcarriers, and calculate the common error estimate based on the reliable subcarriers.

[0119] Step 157: Based on the common error estimation, perform common phase correction and common gain correction on the equalized frequency domain load signal to obtain the corrected frequency domain load signal.

[0120] Step 158: Calculate the error vector amplitude of each subcarrier of each spatial stream based on the corrected frequency domain load signal and the reference frequency domain load signal, and convert the error vector amplitude into the load signal signal-to-noise ratio of each subcarrier of each spatial stream. The time domain load signal is obtained by the signal transmitting end performing a fast Fourier transform on the reference frequency domain load signal.

[0121] In this application, the process of equalizing the frequency domain payload signal based on the channel equalization matrix to obtain the equalized frequency domain payload signal is consistent with the equalization process of the frequency domain training signal. The channel equalization matrix compensates for channel distortion and reduces the impact of noise and interference on the frequency domain payload signal, making the equalized frequency domain payload signal closer to the ideal transmitted signal.

[0122] In this application, a reliable subcarrier refers to a subcarrier with good channel quality and minimal interference. The deviation between its corresponding equalized frequency domain load signal and the known reference signal is mainly caused by common phase error and common gain error, rather than noise or interference. Therefore, it is suitable for calculating common error estimation. The determination of the preset threshold needs to comprehensively consider the actual channel characteristics and evaluation accuracy requirements. The median of the signal-to-interference-plus-noise ratio (SIR) of the training signal corresponding to each spatial stream can be used as the preset threshold. The median reflects the intermediate level of the SIR, avoiding the influence of extreme values ​​and ensuring that the selected reliable subcarriers have good representativeness. For example, if the median SIR of the training signal for spatial stream 1 is 20 dB, then subcarriers with an SIR greater than 20 dB in this spatial stream are selected as reliable subcarriers.

[0123] In this application, common errors mainly include common phase error and common gain error. Common phase error is the phase deviation caused by factors such as frequency offset and channel time variation during signal transmission, while common gain error is the amplitude deviation caused by factors such as inconsistent attenuation and amplifier gain deviation during signal transmission. Using the median calculation can improve the robustness of common error estimation and avoid the influence of abnormal deviations of individual reliable subcarriers on the estimation results.

[0124] In this application, the process of performing common phase correction and common gain correction on the equalized frequency domain load signal based on the common error estimation is to divide the equalized frequency domain load signal by the common error estimation to offset the effects of the common phase error and the common gain error. The corrected frequency domain load signal is closer to the reference frequency domain load signal and can more accurately reflect the influence of noise on the signal.

[0125] In this application, the error vector magnitude (EVM) of each subcarrier in each spatial stream is calculated based on the corrected frequency domain payload signal and the reference frequency domain payload signal. The error vector magnitude is an indicator that measures the degree of deviation between the corrected signal and the ideal reference signal. The larger the deviation, the more serious the impact of noise on the signal and the worse the channel quality.

[0126] Next, this application will describe step 155 or step 158 based on the above embodiments: In this embodiment, the frequency domain load signal is equalized based on the channel equalization matrix to obtain an equalized frequency domain load signal.

[0127] To prevent residual phase error and amplitude contraction caused by minimum mean square error channel equalization from affecting the error vector amplitude and signal-to-noise ratio calculation, the equalized frequency domain load signal needs to be corrected, including common phase correction and common gain correction. Assuming subcarriers... The equalized frequency domain load signal is , It can be represented as:

[0128] For spatial flow Frequency domain symbols In a reliable subcarrier set The frequency domain symbol of the known reference frequency domain load signal is used above. With equalized frequency domain load signal Related calculations and common error estimation ,but It can be represented as:

[0129] Using common error estimation For equalized frequency domain load signals By performing joint correction of common phase and common gain, the corrected frequency domain load signal is obtained. , can be represented as:

[0130] In this embodiment, reliable subcarriers are screened to improve the accuracy of common error estimation. The set of reliable subcarriers can be determined by filtering the signal-to-interference-plus-noise ratio (SINR) of the training signal, selecting subcarriers with high SINR as reliable subcarriers. Assume the training signal-to-interference-plus-noise ratio threshold is... Then the reliable subcarrier set It can be represented as:

[0131] In this embodiment, the corrected frequency domain load signal is used. Frequency domain symbol of the known reference frequency domain load signal The error vector magnitude EVM for each subcarrier in each spatial stream can be expressed as:

[0132] The signal-to-noise ratio of the payload signal for each subcarrier in each spatial stream is obtained by converting the error vector magnitude. , can be represented as:

[0133] Based on steps 155 to 158 above, equalization processing of the frequency domain load signal using the channel equalization matrix can effectively compensate for channel distortion, providing a high-quality signal foundation for subsequent error correction and signal-to-noise ratio (SNR) calculation. Selecting reliable subcarriers and calculating common error estimates based on them ensures the accuracy and robustness of the common error estimates. Reliable subcarriers have better channel quality, and their signal deviations are mainly caused by common errors, accurately reflecting the actual situation of common phase and gain errors. Common phase correction and common gain correction based on common error estimates can successfully offset the influence of common errors on the signal, making the corrected frequency domain load signal closer to the ideal reference signal and more accurately reflecting the impact of noise on channel quality. The calculation of the error vector amplitude quantifies the deviation between the corrected signal and the reference signal. The SNR of the load signal calculated based on this deviation can accurately measure the channel quality of each subcarrier in each spatial flow, complementing the SNR of the training signal. The entire process fully considers various errors and interference factors in the power line channel, ensuring the accuracy and reliability of the load signal SNR and providing a high-quality evaluation index for subsequent weighted fusion. Indicates complex conjugation.

[0134] Continue to refer to Figure 1 In step 160, the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal are weighted and fused to obtain the fused SNR used to evaluate the channel quality of each subcarrier in each spatial stream.

[0135] In this application, a weighted fusion of the training signal signal-to-interference-plus-noise ratio (SNR) and the payload signal SNR is performed to combine the advantages of both evaluation results, yielding a more comprehensive and accurate channel quality assessment index. The training signal SNR is calculated based on a dedicated training signal, providing strong relevance and stability, but it may be affected by the processing of anomalous symbols. The payload signal SNR is calculated based on the actual transmitted payload signal, more closely reflecting the channel quality of real-world communication scenarios, but it may be indirectly affected by factors such as signal modulation and coding. Weighted fusion balances the advantages and disadvantages of the two evaluation results, enabling the final fused SNR to more accurately and robustly reflect the channel quality of each spatial stream and each subcarrier.

[0136] In such Figure 1 In step 160, the weighted fusion of the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal can be performed according to steps 161 to 162 as follows: Step 161: Determine the fusion weight based on the number of frequency domain symbols in the effective training signal or the proportion of abnormal frequency domain symbols, wherein the fusion weight is limited between a first set weight and a second set weight.

[0137] Step 162: Perform weighted fusion of the training signal signal-to-interference-plus-noise ratio (SNR) and the payload signal SNR according to the fusion weights.

[0138] In this application, the fusion weights are determined based on the number of frequency domain symbols or the proportion of anomalous frequency domain symbols in the effective training signal. This is because the number of frequency domain symbols or the proportion of anomalous symbols in the effective training signal reflects the reliability of the signal-to-interference-plus-noise ratio (SNR) of the training signal. A higher number of frequency domain symbols or a lower proportion of anomalous symbols in the effective training signal indicates a more reliable estimate of the SNR, and therefore, it should have a higher weight in the fusion process. Conversely, a lower number of frequency domain symbols or a higher proportion of anomalous symbols in the effective training signal indicates a lower reliability estimate of the SNR, and its weight should be appropriately reduced, while the weight of the payload signal's SNR should be increased accordingly.

[0139] In this application, the fusion weight is limited to a first set weight and a second set weight. The first set weight is the minimum weight, and the second set weight is the maximum weight. Setting this weight range is to avoid excessively high or low weights for any one signal-to-noise ratio (SNR), which could lead to a fusion result biased towards a single evaluation metric and lose the meaning of fusion. The determination of the weight range is based on extensive experiments and real-world scenario testing. The first set weight can be set to 0.2, and the second set weight to 0.8. This range achieves a good balance between the two SNRs, ensuring the robustness and accuracy of the fusion result. For example, when the number of frequency domain symbols in the effective training signal is large (e.g., accounting for more than 90% of the total number of symbols) and the proportion of anomalous symbols is extremely low (e.g., less than 10%), the weight of the training signal SNR can be set to 0.8, and the weight of the payload signal SNR can be 0.2. When the number of frequency domain symbols in the effective training signal is small (e.g., less than 50% of the total number of symbols) and the proportion of abnormal symbols is extremely high (e.g., more than 50%), the weight of the signal-to-interference-plus-noise ratio (SNR) of the training signal can be set to 0.2, and the weight of the SNR of the payload signal can be 0.8. When the number of effective symbols or the proportion of abnormal symbols is at an intermediate level, the weight can be linearly adjusted between 0.2 and 0.8.

[0140] In this application, the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal are weighted and fused according to the fusion weights. This approach combines the advantages of both SNRs. The training signal SNR is calculated based on a specific training signal, exhibiting strong relevance and stability. The payload signal SNR is calculated based on the actual payload signal, more closely reflecting real-world communication scenarios. The fusion of these two factors allows the final fused SNR to more comprehensively and accurately reflect channel quality.

[0141] Next, this application will describe step 161 or step 162 based on the above embodiments: In this embodiment, the fusion weight is determined based on the number of effective frequency domain symbols or the proportion of abnormal frequency domain symbols, and is limited to a preset range. For spatial flow... Subcarriers on Assuming the signal-to-noise ratio after fusion is The fusion weight is ,but It can be represented as:

[0142] Among them, weight , The minimum weight value, The maximum weight can be determined based on the number of effective frequency domain symbols. or abnormal frequency domain symbol count percentage Determine weights .

[0143] Based on steps 161 to 162 above, determining the fusion weights based on the number of frequency domain symbols of the effective training signal or the proportion of abnormal frequency domain symbols allows the fusion weights to adaptively reflect the reliability of the training signal-to-interference-plus-noise ratio (SNR). Dynamic adjustment of the weights ensures the rationality of the fusion result. Limiting the fusion weights within a set range avoids a single evaluation metric dominating the fusion result, ensuring that the advantages of both SNRs are fully utilized. The weighted fusion method combines the stability of the training signal SNR with the practicality of the payload signal SNR, resulting in a fused SNR that more comprehensively, accurately, and robustly reflects the channel quality of each spatial stream and each subcarrier. Compared to a single SNR or SNR evaluation, the fused SNR has higher reliability and adaptability, effectively addressing the complex characteristics and interference variations of power line channels. This fusion process provides the final channel quality evaluation metric for the entire channel evaluation scheme, providing a crucial basis for the communication system to select optimal transmission parameters and configure spatial stream and subcarrier resources, thereby improving the data transmission reliability and communication efficiency of the power line carrier communication system.

[0144] To better illustrate this application, the following will be combined with... Figure 2 To illustrate with a specific embodiment, such as Figure 2 A detailed flowchart of a channel evaluation method for power line carrier communication according to an embodiment of this application is shown, as follows: Figure 2 The method shown can be applied to the physical layer of the next-generation low-voltage power line broadband carrier communication protocol. The system architecture used can be referenced in the "Draft of Key Technologies for Physical Layer of Low-Voltage Power Line Broadband Carrier Communication v2.0".

[0145] like Figure 2 As shown, the specific steps are as follows: 1. Acquisition of frequency domain training signal and frequency domain payload signal: The signal transmitter generates and transmits a signal containing the training signal. and load signal The channel evaluation frame is filled with a known pseudo-random BPSK sequence as the payload signal, and different BPSK modulation symbols are superimposed on the frequency domain symbols of different spatial streams. The signal receiver receives the time-domain training signal and the time-domain payload signal, and converts them into the received frequency-domain training signal through FFT. and received frequency domain payload signal .

[0146] 2. First channel matrix calculation: via and For each subcarrier Least squares estimation is performed, and channel estimation and denoising of multiple frequency domain symbols in the training signal are combined to obtain the first channel matrix. .

[0147] 3. Detection of anomalous frequency domain symbols based on training signal residuals: According to Calculate the residual of each frequency domain symbol in the training signal. Then, the impulse noise detection statistic for each frequency domain symbol in the training signal is calculated. ,calculate The median yields the estimated robust central quantity. ,calculate The median is used to estimate the robust scalar quantity. By comparison and For each frequency domain symbol in the training signal, a decision is made to identify the set of abnormal frequency domain symbols in the training signal. .

[0148] 4. Re-estimation of the second channel matrix and noise covariance matrix based on effective frequency domain symbols: Abnormal frequency domain symbols are removed or weighted accordingly. Weighting the frequency domain symbols yields the effective set of frequency domain symbols. Based on the effective received frequency domain training signal and effectively transmit frequency domain training signals Calculate the second channel matrix ,based on The residual of the training signal is recalculated, which is the noise covariance matrix. .

[0149] 5. Calculate the signal-to-interference-plus-noise ratio (SIR) of the training signal based on the channel equalization matrix: According to the second channel matrix... and noise covariance matrix Construct the minimum mean square error channel equalization matrix for each subcarrier. ,use and Related calculation of equilibrium correlation matrix ,according to Calculate the expected power of the training signal for each spatial stream. and residual crosstalk power ,use and Calculate each spatial flow Then, the post-equalized signal-to-interference-plus-noise ratio (SIR) for each spatial stream and each subcarrier is calculated, i.e. .

[0150] 6. Frequency domain symbol channel equalization and common error correction of the payload signal: utilizing... right Perform channel equalization to obtain the channel equalization output of the frequency domain symbols in each payload signal. This will be greater than the signal-to-interference-plus-noise ratio threshold of the training signal. of The subcarriers are used as a set of reliable subcarriers. ,exist The above utilizes the known frequency domain symbols in the transmitted frequency domain payload signal. and Related calculations and common error estimation ,use right By performing joint correction of common phase and common gain, the corrected load signal channel equalization output is obtained. .

[0151] 7. Calculation of the symbol error vector amplitude and signal-to-noise ratio in the frequency domain of the load signal: using... and Calculate the error vector magnitude for each spatial stream and each subcarrier, i.e., the error vector magnitude. , the magnitude of the error vector Converted to signal-to-noise ratio .

[0152] 8. Weighted fusion of training signal SNR and payload signal SNR: and The signal-to-noise ratio (SNR) is obtained by weighted fusion, i.e. Based on the effective frequency domain symbol count or abnormal frequency domain symbol count proportion Determine the fusion weights , No more than the minimum weight and maximum weight scope.

[0153] Based on the technical solution proposed in this application, by introducing multiple spatial streams to transmit training and payload signals in parallel, and combining this with the differentiated design of pseudo-random binary phase-shift keying modulation symbols, the efficiency of signal transmission and anti-interference capability can be improved. By identifying and processing anomalous frequency domain symbols, the impact of impulse noise, strong interference, and other factors on channel evaluation can be reduced, making the estimation of the second channel matrix and noise covariance matrix more accurate. By equalizing the signal using the optimized channel equalization matrix, channel distortion can be effectively compensated, and signal quality improved. By weighted fusion of the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal, the limitations of a single evaluation method can be avoided. The final fused SNR can comprehensively, accurately, and robustly reflect the channel quality of each spatial stream and each subcarrier, providing a reliable basis for the communication system to select optimal transmission parameters and configure spatial stream and subcarrier resources.

[0154] As another embodiment of this application, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods described in the above embodiments.

[0155] As another embodiment of this application, a computer-readable storage medium is also provided. This computer-readable storage medium may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the methods described in the above embodiments.

[0156] Based on the same inventive concept, embodiments of this application also provide an electronic device. (Reference) Figure 3 The diagram illustrates a structural schematic of a computer system suitable for implementing an electronic device according to embodiments of the present application. The electronic device includes one or more memories 304, one or more processors 302, and at least one computer program (program code) stored in the memories 304 and executable on the processors 302. When the processors 302 execute the computer program, they implement the methods described above.

[0157] Among them, Figure 3 In this document, a bus architecture (represented by bus 300) is used. Bus 300 may include any number of interconnected buses and bridges, linking various circuits including one or more processors represented by processor 302 and memory represented by memory 304. Bus 300 may also link various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. Receiver 301 and transmitter 303 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 302 is responsible for managing bus 300 and general processing, while memory 304 can be used to store data used by processor 302 during operation.

[0158] The functions described herein may be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions may be stored as one or more instructions or codes on or transmitted via a computer-readable medium. Other examples and embodiments are within the scope and spirit of this application and the appended claims. For example, due to the nature of software, the functions described above may be implemented using software executed by a processor, hardware, firmware, hardwired, or any combination thereof. Furthermore, the functional units may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit.

[0159] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0160] The units described as separate components may or may not be physically separate. Similarly, the components of the control device may or may not be physical units; they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0161] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium, including instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0162] The above description is merely an embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A channel evaluation method for power line carrier communication, characterized in that, The method includes: Acquire a channel evaluation frame, which includes a time-domain training signal and a time-domain payload signal transmitted by the signal transmitter through multiple spatial streams; The time-domain training signal and the time-domain payload signal are subjected to Fast Fourier Transform to obtain the frequency-domain training signal and the frequency-domain payload signal, respectively. The frequency-domain training signal corresponding to each spatial stream includes multiple repeated frequency-domain symbols, and different pseudo-random binary phase-shift keying modulation symbols are superimposed on the frequency-domain symbols of different spatial streams. The frequency-domain payload signal corresponding to each spatial stream is filled with a known pseudo-random binary phase-shift keying sequence. Both the frequency-domain training signal and the frequency-domain payload signal are composed of multiple subcarriers. Estimate the first channel matrix of the frequency domain training signal, and calculate the training signal residual based on the first channel matrix, so as to identify abnormal frequency domain symbols in the frequency domain training signal according to the training signal residual; The abnormal frequency domain symbols in the frequency domain training signal are processed to obtain the effective training signal, and the second channel matrix and noise covariance matrix of the effective training signal are estimated. A channel equalization matrix is ​​constructed based on the second channel matrix and the noise covariance matrix. The frequency domain training signal and the frequency domain payload signal are then processed based on the channel equalization matrix to obtain the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal for each subcarrier of each spatial stream. The signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal are weighted and fused to obtain the fused SNR used to evaluate the channel quality of each subcarrier in each spatial stream.

2. The method according to claim 1, characterized in that, The estimation of the first channel matrix of the frequency domain training signal includes: Based on the known training signal stored locally, the least squares estimation method is used to perform channel estimation and denoising on multiple frequency domain symbols in the frequency domain training signal to obtain the first channel matrix of the frequency domain training signal.

3. The method according to claim 2, characterized in that, The calculation of the training signal residual based on the first channel matrix includes: Based on the first channel matrix and the known training signal, determine the training signal estimates corresponding to multiple frequency domain symbols; Calculate the deviation between each frequency domain symbol in the frequency domain training signal and the training signal estimate corresponding to each frequency domain symbol to obtain the training signal residual corresponding to each frequency domain symbol.

4. The method according to claim 3, characterized in that, The step of identifying anomalous frequency domain symbols in the frequency domain training signal based on the training signal residual includes: Based on the training signal residuals, an impulse noise detection statistic corresponding to each frequency domain symbol is constructed, and the robust central quantity and robust scale quantity of the impulse noise detection statistic are robustly estimated by the median. The impact noise detection statistic is compared with the robust center quantity, the robust scale quantity, and the preset threshold coefficient; If the absolute value of the difference between the impact noise detection statistic and the robustness center quantity is greater than the product of the preset threshold coefficient and the robustness scale quantity, then each frequency domain symbol is determined to be an abnormal frequency domain symbol.

5. The method according to claim 4, characterized in that, The processing of abnormal frequency domain symbols in the frequency domain training signal includes: Remove the abnormal frequency domain symbols from the frequency domain training signal; or... Weights are assigned to each frequency domain symbol in the frequency domain training signal, wherein the weights assigned to abnormal frequency domain symbols are lower than the weights assigned to normal frequency domain symbols.

6. The method according to claim 5, characterized in that, The estimation of the second channel matrix and noise covariance matrix of the effective training signal includes: The known training signal is processed according to the processing method for abnormal frequency domain symbols in the frequency domain training signal to obtain a valid known training signal; Based on the effective known training signal, the least squares estimation method is used to perform channel estimation and denoising on multiple frequency domain symbols in the effective training signal to obtain the second channel matrix of the effective training signal. The noise covariance matrix is ​​obtained by recalculating the training signal residual based on the effective training signal; or, the noise covariance matrix is ​​obtained by calculating the training signal residual in a weighted manner according to the weights of the effective training signal.

7. The method according to claim 6, characterized in that, The channel equalization matrix is ​​a minimum mean square error channel equalization matrix. Based on the channel equalization matrix, the frequency domain training signal is processed to obtain the signal-to-interference-plus-noise ratio (SIR) of the training signal for each subcarrier in each spatial stream, including: Calculate the equalization correlation matrix based on the second channel matrix and the channel equalization matrix; Based on the equalization correlation matrix, calculate the expected power of each subcarrier of each spatial stream and the residual crosstalk power of each subcarrier of other spatial streams in the frequency domain training signal; The noise power of each subcarrier in each spatial stream is calculated by combining the channel equalization matrix and the noise covariance matrix. The training signal-to-interference-plus-noise ratio (SNR) for each subcarrier in each spatial stream is calculated based on the expected power, the residual crosstalk power, and the noise power.

8. The method according to claim 7, characterized in that, The frequency domain payload signal is processed based on the channel equalization matrix to obtain the signal-to-noise ratio of the payload signal for each subcarrier in each spatial stream, including: The frequency domain load signal is equalized based on the channel equalization matrix to obtain an equalized frequency domain load signal. Subcarriers with a signal-to-interference-plus-noise ratio (SINR) greater than a preset threshold are selected as reliable subcarriers, and a common error estimate is calculated based on the reliable subcarriers. Based on the common error estimation, the equalized frequency domain load signal is subjected to common phase correction and common gain correction to obtain the corrected frequency domain load signal; The error vector amplitude of each subcarrier in each spatial stream is calculated based on the corrected frequency domain payload signal and the reference frequency domain payload signal, and the error vector amplitude is converted into the signal-to-noise ratio of the payload signal of each subcarrier in each spatial stream. The time domain payload signal is obtained by the signal transmitting end performing a fast Fourier transform on the reference frequency domain payload signal.

9. The method according to claim 8, characterized in that, The weighted fusion of the signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal includes: The fusion weight is determined based on the number of frequency domain symbols in the effective training signal or the proportion of abnormal frequency domain symbols, and the fusion weight is limited between a first set weight and a second set weight. The signal-to-interference-plus-noise ratio (SNR) of the training signal and the SNR of the payload signal are weighted and fused according to the fusion weights.

10. An electronic device, characterized in that, The electronic device includes one or more processors and one or more memories, wherein at least one piece of program code is stored in the one or more memories, and the at least one piece of program code is loaded and executed by the one or more processors to implement the method as claimed in any one of claims 1 to 9.

Images