Self-adaptive channel estimation method and device for power line OFDM (Orthogonal Frequency Division Multiplexing) communication system
By using an adaptive channel estimation method, the filtering configuration of the channel estimation method is dynamically adjusted according to the power line channel state, which solves the problems of wasted computational resources and insufficient accuracy in fixed structure channel estimation in power line OFDM communication systems, and achieves a balance between computational complexity and estimation accuracy under different channel conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-12
- Publication Date
- 2026-04-07
AI Technical Summary
In existing power line OFDM communication systems, fixed-structure channel estimation methods cannot balance estimation accuracy and computational efficiency in changing channel environments, resulting in wasted computational resources in good channels and insufficient filtering performance in poor channels.
By acquiring the initial frequency domain response of the channel, calculating the parameters of the channel frequency selective fading intensity and noise level, dynamically querying the configuration lookup table, and adjusting the filtering configuration of the channel estimation method, including the application interval and the filtering order, adaptive channel estimation is achieved by combining high-complexity and low-complexity channel estimation methods.
Under different channel conditions, the channel estimation method is dynamically adjusted to reduce computational complexity, save hardware resources, and ensure channel estimation accuracy and communication reliability, thereby achieving synergistic optimization of computational complexity and estimation accuracy.
Smart Images
Figure CN121814512A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power line communication technology, and in particular to an adaptive channel estimation method, receiving device, and computer storage medium for power line orthogonal frequency division multiplexing (OFDM) communication systems. Background Technology
[0002] Power line communication (PLC) is a communication technology that utilizes existing power line networks for signal transmission. Its core advantage lies in eliminating the need to lay dedicated communication lines, thus significantly reducing deployment and maintenance costs. However, power lines were originally designed for power transmission, and using them as a communication medium presents challenges due to their inherent characteristics, including high noise, strong attenuation, and impedance variations. To address these challenges, Orthogonal Frequency Division Multiplexing (OFDM) technology has become crucial for realizing broadband power line communication.
[0003] In power line OFDM communication systems, channel estimation is a crucial step in achieving reliable demodulation. Due to severe frequency-selective fading, multipath effects, and impulse noise in power line channels, signals experience significant distortion and loss of quality during transmission. Accurate channel estimation allows the system to acquire channel state information for each subcarrier, providing necessary compensation coefficients for subsequent frequency domain equalization.
[0004] Currently, some channel estimation methods with strong noise reduction capabilities, such as linear minimum mean square error (LMMSE) filtering or transform domain filtering, while offering high estimation accuracy, also have high computational complexity, placing significant pressure on the hardware implementation's processing timing. Furthermore, for ease of control, existing implementations often employ fixed filtering structures, whose computational load is fixed and does not change with channel conditions. This leads to wasted computational resources under favorable channel conditions, while under adverse conditions, the fixed computational load may fail to provide sufficient filtering performance. Therefore, there is an urgent need for a channel estimation scheme that can dynamically adjust based on real-time channel conditions, minimizing computational complexity while maintaining estimation accuracy. Summary of the Invention
[0005] The purpose of this invention is to solve the technical problem that the fixed-structure channel estimation method in the prior art cannot balance estimation accuracy and computational efficiency in the changing power line channel environment, and to propose an adaptive channel estimation method and device for power line OFDM communication systems.
[0006] The technical problem of this invention is solved by the following technical solution: In a first aspect, the present invention provides an adaptive channel estimation method for a power line OFDM communication system, comprising the following steps: S1. Obtain the initial frequency domain response estimate of the current channel; S2. Calculate the first channel parameter to characterize the channel frequency selective fading intensity based on the initial frequency domain response estimate; S3. Calculate the second channel parameter used to characterize the channel noise level based on the initial frequency domain response estimate; S4. Based on the first channel parameters and the second channel parameters, query the preset configuration lookup table to dynamically determine the filtering configuration of the first type of channel estimation method. The filtering configuration includes the application interval and filtering order of the first type of channel estimation method. The configuration lookup table records the filtering configuration that minimizes the total computational complexity under different channel states while meeting the preset channel estimation performance threshold. S5. Using the subcarriers determined by the application interval as nodes, a first type of channel estimation method with first filtering capability is used and the filtering order is used to perform channel estimation on the corresponding node subcarriers to obtain the node subcarrier channel estimation value. S6. For non-node subcarriers, a second type of channel estimation method with lower computational complexity than the first type of channel estimation method is used, and channel estimation is performed based on the channel estimation values of the adjacent node subcarriers.
[0007] In some embodiments, the following technical features are also included: Step S1 specifically includes the following steps: S11. Receive the time-domain signal of the preamble sequence PREAM or training sequence TF in the power line communication system. S12. Perform a fast Fourier transform on the time-domain signal to obtain the frequency-domain received data; S13. Divide the frequency domain received data by the known frequency domain transmitted data to obtain the least squares estimate on each subcarrier, which is used as the initial frequency domain response estimate.
[0008] In some embodiments, before performing a fast Fourier transform on the time-domain signal in step S12, the following steps are further included: If the PREAM or TF is repeatedly transmitted by multiple identical OFDM symbols, the time-domain signals of the multiple symbols are first averaged in the time domain, and then a fast Fourier transform is performed.
[0009] In some embodiments, in step S2, calculating the first channel parameter used to characterize the channel frequency-selective fading intensity specifically includes the following steps: S21. Perform a fast inverse Fourier transform on the initial frequency domain response estimate to obtain the time-delay domain channel impulse response; S22. Calculate the power delay spectrum based on the time-delay domain channel impulse response; S23. Identify and filter out effective paths from the power delay spectrum to obtain a power delay spectrum containing only effective paths; S24. Calculate the root mean square delay spread based on the power delay spectrum containing only the effective path, and use it as the first channel parameter.
[0010] In some embodiments, the power line OFDM communication system is a multiple-input multiple-output system, and step S22 specifically includes the following steps: Calculate the power delay spectrum between each transmit and receive antenna pair; In step S23, identifying and filtering effective paths from the power delay spectrum specifically includes the following steps: The power delay spectrum of each transmit and receive antenna pair is averaged to obtain the average power delay spectrum, and the effective path is identified from the average power delay spectrum.
[0011] In some embodiments, step S2 specifically includes the following steps: Based on the initial frequency domain response estimate, the frequency domain amplitude crossover rate is calculated and used as the first channel parameter.
[0012] In some embodiments, step S3, calculating the second channel parameter used to characterize the channel noise level, specifically includes the following steps: S31. Divide the channel estimation bandwidth into multiple sub-bandwidths; S32. Calculate the signal-to-noise ratio estimate for each sub-bandwidth and use it as the second channel parameter corresponding to each sub-bandwidth. Accordingly, the query of the preset configuration lookup table and the subsequent channel estimation steps are executed independently for each sub-bandwidth.
[0013] In some embodiments, the first type of channel estimation method is linear minimum mean square error filtering or transform domain filtering; the second type of channel estimation method is linear interpolation, cubic spline interpolation or Gaussian interpolation.
[0014] In a second aspect, the present invention provides a power line orthogonal frequency division multiplexing (OFDM) communication receiving device, comprising: one or more processors; a memory; and one or more computer programs; wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and the one or more computer programs include instructions for performing the above-described method.
[0015] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described method.
[0016] The beneficial effects of this invention compared to the prior art include: The adaptive channel estimation method provided by this invention dynamically determines the optimal filtering configuration (including application interval and filtering order) for the first type of channel estimation method by calculating a first channel parameter characterizing the intensity of channel frequency-selective fading and a second channel parameter characterizing the channel noise level in real time, and then querying a pre-set configuration lookup table based on these parameters. This technical solution enables the system to adaptively adjust the usage strategy of high-complexity, high-precision algorithms (first type of method) and low-complexity, low-precision algorithms (second type of method) according to the real-time channel status. Its advantages are that, under good channel conditions with high signal-to-noise ratio or weak channel frequency selectivity, the system significantly reduces overall computational complexity, saves hardware resources, and reduces power consumption by increasing the application interval of the first type of method and / or decreasing its filtering order, and relying more on the second type of method for interpolation estimation. Conversely, under poor channel conditions with low signal-to-noise ratio or strong channel frequency selectivity, the system ensures sufficiently accurate channel estimates on key node subcarriers by decreasing the application interval of the first type of method and / or increasing its filtering order, thereby guaranteeing the overall accuracy of channel estimation and the reliability of the communication system. Ultimately, this method achieves synergistic optimization of the conflicting objectives of computational complexity and estimation accuracy under different channel conditions, maximizing the reduction of computational load without sacrificing system performance.
[0017] In addition, some embodiments also have the following beneficial effects: By dividing the estimated bandwidth into multiple sub-bandwidths and performing channel estimation independently, the challenge of uneven noise power spectral density in power line channels can be addressed, further improving the accuracy of the adaptive effect.
[0018] By using root mean square delay spread or frequency domain amplitude crossover rate as the first channel parameter, a reliable and feasible quantitative means is provided for accurately evaluating channel frequency selectivity.
[0019] By encompassing various first-type methods such as LMMSE and transform domain filtering, as well as various second-type methods such as linear interpolation and spline interpolation, the solution provides flexible options for different implementation scenarios, enhancing its versatility.
[0020] Other beneficial effects of the embodiments of the present invention will be further described below. Attached Figure Description
[0021] Figure 1 This is a flowchart of an adaptive channel estimation method for a power line OFDM communication system in an embodiment of the present invention.
[0022] Figure 2This is a basic flowchart of adaptive dynamic filtering for power line communication systems in this embodiment of the invention.
[0023] Figure 3a This is a schematic diagram of a scenario with strong frequency domain correlation in an embodiment of the present invention.
[0024] Figure 3b This is a schematic diagram of a scenario with weak frequency domain correlation in an embodiment of the present invention.
[0025] Figure 4 This is a schematic diagram of channel estimation in an embodiment of the present invention, which segments the channel based on the channel state and uses different channel estimation structures for each segment.
[0026] Figure 5 This diagram illustrates a comparison of frame error rates between the traditional fixed LMMSE filtering method and the filtering method in this embodiment of the invention under an additive white Gaussian noise channel.
[0027] Figure 6 This diagram illustrates a comparison of frame error rates between the traditional fixed LMMSE filtering method and the filtering method in this embodiment of the invention under multipath channels. Detailed Implementation
[0028] The present invention will be further described below with reference to the accompanying drawings and preferred embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.
[0029] It should be noted that the directional terms such as left, right, up, down, top, and bottom used in this embodiment are only relative concepts or are based on the normal use of the product, and should not be considered as restrictive.
[0030] Implementation concept: The core of this invention lies in dynamically adjusting the usage strategy (i.e., filtering configuration, including the application interval and filtering order of the first type of algorithm) for channel estimation based on the real-time state of the power line channel (frequency-selective fading intensity and noise level), to achieve an optimal balance between computational complexity and estimation accuracy. Its technical principle is to quantify channel state parameters and, based on a pre-established configuration lookup table built through extensive simulations with the goal of satisfying performance thresholds with optimal complexity, guide real-time algorithm selection and parameter configuration.
[0031] Example 1 This invention provides an adaptive channel estimation method for power line OFDM communication systems. It dynamically adjusts the structure of the channel estimation filter based on the current channel state to minimize the computational load of the module without sacrificing channel estimation performance. The channel estimation filter structure involves a more accurate but computationally intensive first-type channel estimation method (e.g., minimum mean squared error (MMSE) or transform domain filtering) and a less computationally intensive second-type channel estimation method (e.g., linear interpolation or cubic spline interpolation). The method involves dynamically adjusting the subcarrier spacing using the computationally intensive first-type channel estimation method by referencing relevant parameters directly related to the power line communication system channel. The remaining subcarriers use the computationally less intensive second-type channel estimation method. This method considers the differences between various power line channels and minimizes the computational load of the module without sacrificing filtering performance based on specific channel conditions.
[0032] The overall process of the method in this embodiment is as follows: Figure 1 As shown, it includes the following steps: S1. Obtain the initial frequency domain response estimate of the current channel; specifically, use the frequency domain transform (Fast Fourier Transform) of the preamble sequence (PREA) or training field (TF) symbols in the power line communication system to perform least squares estimation to obtain the coarse channel estimate result of the current channel. Step S1 specifically includes the following steps: S11. Receive the time-domain signal of the preamble sequence PREAM or training sequence TF in the power line communication system. S12. Perform a fast Fourier transform on the time-domain signal to obtain the frequency-domain received data; S13. Divide the received frequency domain data by the known transmitted frequency domain data to obtain the least squares estimate on each subcarrier, which is used as the initial frequency domain response estimate. Specifically, let... For PREAM or TF frequency domain Data received in the frequency domain of each subcarrier. For the known first If the frequency domain of the nth subcarrier transmits data, then the nth... Least Squares (LS) estimation of subcarriers for
[0033] Since the PREM or training sequence of the power line protocol is usually repeatedly transmitted in multiple identical OFDM symbols, in order to improve noise reduction capability, the PREM or TF can be averaged in the time domain before performing frequency domain transformation.
[0034] S2. Based on the initial frequency domain response estimate, calculate the first channel parameter used to characterize the channel frequency-selective fading intensity; step S2 specifically includes the following steps: S21. Perform an inverse fast Fourier transform on the initial frequency domain response estimate to obtain the time-delay domain channel impulse response. Specifically, let... For all available LS results, Given the number of subcarriers, the channel response in the time-delay domain is obtained by performing an inverse Fast Fourier Transform (IFFT) on the least squares result of all subcarriers.
[0035] in To estimate the obtained noisy channel response, The inverse Fourier transform size is... This is the noise-free channel response. It is noise.
[0036] S22. Calculate the power delay spectrum based on the time-delay domain channel impulse response. Specifically, after obtaining the time-delay domain channel response... Then, using The power delay profile (PDP) is calculated as follows:
[0037] For simplicity of symbol description, the above discussion only describes the single-input single-output (SISO) scenario. In the MIMO (Multiple-Input Multiple-Output) scenario, the power line OFDM communication system is a multiple-input multiple-output system, and step S22 specifically includes the following steps: Calculate the power delay spectrum between each transmit and receive antenna pair; S23. Identify and filter effective paths from the power delay spectrum to obtain a power delay spectrum containing only effective paths; in step S23, identifying and filtering effective paths from the power delay spectrum specifically includes the following steps: The power delay spectrum of each transmit / receive antenna pair is averaged to obtain the average power delay spectrum, and the effective path is identified from the average power delay spectrum. The average power delay spectrum is:
[0038] in For sending port and receiving end The corresponding power delay spectrum of the first Sample points.
[0039] According to the power delay spectrum We can select the noise path and the effective path, and denote the power delay spectrum containing only the effective path as... The selection of effective paths can be achieved by: calculating a global noise power reference for the power delay spectrum (PDP) (e.g., through noise estimation at the frequency domain pilot or PDP tail averaging), and identifying paths with power 3-5 dB higher than this reference (or other thresholds determined by simulation) as effective paths. The identification of effective paths can be achieved through one of the following methods: (a) Calculate the global noise power reference for the power delay spectrum, and determine the effective path as the path whose power is consistently higher than a certain threshold (e.g., 3-5 dB) of the reference. (b) Adopt an adaptive threshold based on signal-to-noise ratio. For example, the threshold can be set as X + Y *σ_n, where σ_n is the standard deviation estimate of the noise power, and X and Y are constants preset according to the system performance requirements. (c) Alternatively, by comparing the power delay spectra of multiple consecutive symbols or frames, persistent paths can be identified as effective paths to suppress noise disturbances.
[0040] S24. Based on the power delay spectrum containing only the effective path, calculate the root mean square (rms) delay spread as the first channel parameter. Specifically, use the power delay spectrum obtained above. Calculate the root mean square delay spread
[0041] The above average delay It can be represented as
[0042] Note that the unit of the above root mean square delay spread is the delay domain resolution of the power delay spectrum, i.e. ( );in The bandwidth of the reference pilot used to calculate the power delay spectrum.
[0043] S3. Based on the initial frequency domain response estimate, calculate the second channel parameter used to characterize the channel noise level. Specifically, calculate the SNR (Signal-to-Noise Ratio) value in the frequency domain. Since the pilots of the power line communication protocol are closely arranged in the frequency domain, the results of LS are used. Calculate the estimated value of SNR in the frequency domain.
[0044] S4. Based on the first channel parameters and the second channel parameters, query a preset configuration lookup table to dynamically determine the filtering configuration of the first type of channel estimation method. The filtering configuration includes the application interval and filtering order of the first type of channel estimation method. The configuration lookup table records the filtering configuration that minimizes the total computational complexity under different channel states while meeting the preset channel estimation performance threshold.
[0045] Specifically, based on the root mean square delay spread and SNR estimate calculated in the above steps, the filtering structure (filter node spacing of the first type of channel estimation method (e.g., a filter node spacing of 2 means that the first type of channel estimation method is used for channel estimation every two subcarriers), filter order, and other information) is determined according to the lookup table. Generally speaking, a larger root mean square delay spread indicates a stronger frequency selectivity of the channel, and the filter spacing of the first type of channel estimation method should be smaller; a smaller SNR estimate indicates a stronger filtering capability of the first type of channel estimation method is required, and therefore the filter order should be larger. The principle for constructing the lookup table is to minimize the total computational complexity while ensuring that the mean square error of channel estimation (MSE) is lower than the threshold required by the system. Through extensive simulations under different channel models (such as the IEEE 1901 channel model) and different SNRs, the optimal filtering configuration (node spacing and order) corresponding to each channel state that meets the performance requirements is statistically determined, thereby forming the lookup table.
[0046] S5. Using the subcarriers determined by the application interval as nodes, a first type of channel estimation method with first filtering capability is used and the filtering order is used to perform channel estimation on the corresponding node subcarriers to obtain the node subcarrier channel estimation value.
[0047] Specifically, based on the filtering structure calculated above, channel estimation is performed on some subcarriers using the first type of channel estimation method.
[0048] S6. For non-node subcarriers, a second type of channel estimation method with lower computational complexity than the first type of channel estimation method is used, and channel estimation is performed based on the channel estimation values of the adjacent node subcarriers.
[0049] Specifically, channel estimation is performed on the remaining subcarriers based on the results of step S5 and the second type of channel estimation method.
[0050] Step S3, calculating the second channel parameter used to characterize the channel noise level, specifically includes the following steps: S31. Divide the channel estimation bandwidth into multiple sub-bandwidths; specifically, segment the target channel estimation bandwidth and then calculate the corresponding SNR for each segment. This is because, unlike many other communication channels, the noise in power line communication channels cannot typically be described as additive white Gaussian noise (AWGN). Therefore, in broadband power line communication systems, the power spectral density of the noise may differ significantly across different bandwidth ranges, which also means that different noise reduction capabilities may be required within different bandwidth ranges. The size of the segment can be determined by the total bandwidth.
[0051] S32. Calculate the estimated signal-to-noise ratio (SNR) for each of the sub-bandwidths, using it as the second channel parameter corresponding to each sub-bandwidth. Accordingly, the query of the preset configuration lookup table and the subsequent channel estimation steps are executed independently for each sub-bandwidth. Specifically, after calculating the SNR for different bandwidth ranges in step S3, subsequent steps are performed for each bandwidth range.
[0052] In step S2, the root mean square delay spread is used as the metric to determine the strength of frequency domain selectivity. This metric can also be other parameters, such as the frequency-domain level crossing rate (LCRf). The filtering interval for the first type of channel estimation method is determined using a lookup table based on the LCRf per unit bandwidth. This lookup table can be obtained through extensive simulations.
[0053] In step S5, the first type of channel estimation method is characterized by strong filtering capability but may have slightly higher computational complexity. One possible first-type channel estimation method is Linear Minimum Mean Square Error (LMMSE) filtering. Since power line communication transmission is bursty, the time interval between the transmission of the previous frame and the current frame is random. Therefore, if LMMSE filtering is used, the channel information of the current frame needs to be used to obtain the frequency domain LMMSE filter coefficients. Channel estimation based on LMMSE for a subcarrier can be expressed as...
[0054] in This is an estimated value. For size The row vector, The closest to the absolute distance to the subcarrier to be estimated The size of the LS results of each pilot subcarrier is The column vector. Filter coefficients. It can be represented as
[0055] in The subcarrier positions and column vectors to be filtered middle The size of each pilot subcarrier is The correlation vector. column vector middle The size of each pilot subcarrier is The autocorrelation matrix. The correlation can be obtained from the relationship between the correlation function and the Fourier transform of the power delay spectrum. This refers to the linear signal-to-noise ratio (SNR) value calculated according to steps S3 and S4. For example, for a 16th-order LMMSE filter, when the filter interval is 1 (i.e., all subcarriers are filtered using channel estimation mode 1), 16 sets of values of size 1 need to be generated. The coefficients; when the filtering interval is When (i.e., filtering is performed through channel estimation method 1 every other subcarrier), only 8 groups of the same size need to be generated. The coefficients; that is, half of the computation can be saved in the filter coefficient generation stage.
[0056] Another possible first-type channel estimation method is transform domain filtering, which uses the power delay spectrum with only effective paths as described in step S2. The corresponding denoised channel impulse response is obtained. Then through the channel impulse response The DFT is performed to transform the data into the frequency domain response of the corresponding subcarrier according to the appropriate filtering interval. Clearly, by increasing the filtering interval of the first type of channel estimation method, the computational cost of the DFT can be reduced.
[0057] The second type of channel estimation method in step S6 is characterized by its weak filtering capability but extremely low computational complexity. One possible second type of channel estimation method is linear interpolation. Assume... and For two adjacent subcarriers estimated using the first type of channel estimation method, Given the filtering interval for the first type of channel estimation method, then the subcarrier The channel response at that location can be expressed as
[0058] In the above formula and For separation The frequency responses of the two most recent subcarriers obtained using the first type of channel estimation method after filtering.
[0059] The first type of channel estimation method is linear minimum mean square error filtering or transform domain filtering; the second type of channel estimation method is linear interpolation, cubic spline interpolation or Gaussian interpolation, high-intermediate interpolation, etc.
[0060] This invention relates to a channel filtering mechanism for frequency domain subcarriers in a power line communication system. It proposes to dynamically adjust the filtering structure based on the current channel characteristics, and more specifically, the frequency selectivity of the channel and the required actual noise reduction capability. Ultimately, while ensuring the accuracy of channel estimation, the computational load of the channel parameter estimation module is reduced as much as possible.
[0061] The method of this invention can achieve the goal of minimizing the computational load of the module without sacrificing channel estimation performance. Its core innovation lies in combining two channel estimation methods with different noise reduction capabilities and complexities. The first type of channel estimation method has strong noise reduction capabilities but high computational complexity, while the second type of channel estimation method has lower computational complexity but lower computational load. This second type of channel estimation method filters the results obtained from the first type of channel estimation method. The combination of the first and second type of channel estimation methods is dynamically adjusted according to the current channel state to adapt to the current power line communication channel.
[0062] Example 2 This embodiment provides a filtering method that dynamically adjusts the structure of the channel estimation filter based on the current power line communication channel state, such as... Figure 2 As shown, it includes the following steps: Step 1: Use the frequency domain transform of the preamble sequence PREM or training sequence TF symbols in the power line communication system to perform least squares estimation to obtain a coarse channel estimate for the current channel. Since the PREM or training sequence of the power line protocol is generally transmitted repeatedly in multiple identical OFDM symbols, in order to improve noise reduction capability, the PREM or TF can be averaged in the time domain before performing the frequency domain transform.
[0063] Step 2: Perform an inverse fast Fourier transform on the least squares (LS) estimation result obtained in Step 1 to obtain the channel impulse response in the time delay domain and its corresponding power delay spectrum (PDP). If MIMO transmission is involved, the power delay spectrum can be averaged across the transmit and receive antenna pairs to obtain the average power delay spectrum. Use the average power delay spectrum to select the effective path.
[0064] The root mean square delay spread (root mean square delay estimate) is calculated using the power delay spectrum obtained above.
[0065] Another way to implement step two is to use the results of LS to calculate the frequency-domain level crossing rate (LCRf) after smoothing and denoising.
[0066] Step 3: Segment the bandwidth of PREAM or TF and calculate the SNR value of each frequency segment. The segmentation rule can be uniform segmentation or non-uniform segmentation based on prior knowledge of the noise power spectral density. A simple implementation is to divide the total bandwidth into K equal segments, where K is an integer greater than 1, and its specific value can be determined based on a trade-off between the total number of subcarriers and implementation complexity. Figure 4 As shown, based on the specific channel state (empty subcarrier, data subcarrier, data subcarrier), the frequency band for transmitting data is divided into four segments (channel estimation structure 1, channel estimation structure 2, channel estimation structure 3, channel estimation structure 4), and each frequency segment has a different channel estimation structure (i.e., different filtering method 1 filtering interval and / or different filtering order).
[0067] Step 4: Based on the root mean square delay spread / LCRf and SNR estimates calculated in the previous steps, determine the filter structure configuration (filter interval, filter order, etc. for each segment in Step 3) using a lookup table. Generally, a larger root mean square delay spread or a larger LCRf indicates stronger frequency selectivity of the channel, and the filter interval of Channel Estimation Method 1 (Type I channel estimation method) should be smaller; a smaller SNR estimate indicates a stronger filtering capability required for Channel Estimation Method 1, therefore a larger filter order is needed. Figure 3a and Figure 3b This embodiment presents two different frequency domain filtering structures for two different channel scenarios (with varying frequency correlations): Figure 3a This is a schematic diagram of a scenario with strong frequency domain correlation. Figure 3b This diagram illustrates a scenario with weak frequency domain correlation. The specific values for the corresponding lookup table were obtained through extensive simulations. The pre-set configuration lookup table is generated through the following offline process: establishing a typical channel model library and signal-to-noise ratio range covering the target power line communication scenario; for each set of channel parameters (first channel parameter, second channel parameter) with discrete values, traversing all possible filtering configurations (application interval, filtering order); calculating the channel estimation performance (e.g., mean square error MSE) and computational complexity for each configuration through simulation; finally, for each set of channel parameters, selecting the filtering configuration that meets the preset performance threshold (e.g., MSE < target value) and has the lowest computational complexity, and storing it in the lookup table.
[0068] Step 5: Based on the filter structure calculated above, use channel estimation method 1 to perform channel estimation on some subcarriers according to the selected filter interval and order.
[0069] Step 6: Based on the results of Step 5 above and Channel Estimation Method 2 (Second Type of Channel Estimation Method), perform channel estimation on the remaining subcarriers.
[0070] For example, if the State Grid's HPLC (High-Performance Power Line Communication) band 0 (2-12 MHz) is used for data transmission, the 12th-order LMMSE channel estimation method involves a total of This involves complex multiplication and 4932 complex additions. However, in many cases where the power line channel has good frequency flatness or poor frequency selectivity, the LMMSE filter interval is used. The remaining subcarriers, using second-order interpolation, involve a total of Multiplication of complex numbers and This involves multiple complex additions. It can be seen that, in terms of filtering operations, the latter requires more than half the computational cost of the former. Furthermore, for LMMSE filtering, Wiener filter coefficient generation is one of the main computational burdens of early LMMSE filtering. If all subcarriers use LMMSE channel estimation, then in terms of coefficient generation, a total of 1... Inverse of a matrix There are approximately 1728 complex multiplications and 1728 complex additions; if the LMMSE filter interval is changed from 1 to 3, the number of complex multiplications and additions in coefficient generation is reduced to approximately 576.
[0071] The present invention proposes a structure for dynamically adjusting the channel estimation filter based on the current power line communication channel state characteristics, which reduces the computational load of the module as much as possible without sacrificing channel estimation performance, and has the following advantages: (1) By dynamically adjusting the filtering configuration of channel estimation method 1, including the filtering interval and the filtering order, and combining it with the subsequent channel estimation method 2, a balance between channel estimation accuracy and computational load can be achieved. (2) In a high SNR low frequency domain selectivity environment, the filtering interval and order of channel estimation method 1 can be reduced, and the proportion of subcarriers estimated by channel estimation method 2 can be increased to save computation. (3) In a high SNR high frequency domain selectivity environment, the filtering interval of channel estimation method 1 can be increased and the filtering order of channel estimation method 1 can be reduced to achieve a balance between channel estimation performance and computational load. (4) In a low SNR and low frequency selectivity environment, the order of channel estimation method 1 can be increased, the filtering interval of channel estimation method 1 can be reduced, and the proportion of subcarriers estimated by channel estimation method 2 can be increased to achieve a balance between channel estimation performance and computational load. (5) In a low SNR and high frequency selectivity environment, the filtering interval and order of channel estimation method 1 can be increased, and the proportion of subcarriers estimated by channel estimation method 2 can be reduced to improve performance.
[0072] The advantages of the embodiments of the present invention also include: (1) Dynamically change the computational load based on the channel state: In this embodiment, the filtering structure of channel estimation methods 1 and 2 is adjusted based on the power line communication channel state, and the computational load of channel estimation is dynamically adjusted on the basis of satisfying the channel estimation performance. (2) Architecture universality: This embodiment does not depend on a specific frequency, avoiding the development of independent algorithms for each frequency band and reducing deployment complexity; at the same time, the ideas implemented in this embodiment can also be ported to time domain filtering.
[0073] Example 3 This embodiment provides a method for dynamically adjusting the channel estimation filter structure based on the current power line communication channel state characteristics. It considers 411 data subcarriers in the State Grid HPLC band 0 (2-12 MHz) for data transmission. The specific implementation steps are as follows: Step 1: Time-domain data of different symbols of PREAM or TF in power line carrier communication Perform averaging. Then, average the time-domain data. The data is converted into frequency domain data by FFT operation. The least-squares (LS) result of the frequency domain channel response is obtained by dividing the known PREAM or TF frequency domain transmitted data by the frequency domain received data. .
[0074] Step 2: Utilize the least squares (LS) results of the frequency domain response described above. The channel impulse response is obtained by performing IFFT operations. The power delay spectrum was obtained from the channel impulse response. Using the above Obtain the power delay spectrum containing only the effective path. Using the power delay spectrum that only contains the effective path mentioned above The root mean square delay spread was calculated. For example, in this embodiment, the calculated root mean square delay spread It is 60 ns.
[0075] Step 3: Segment the data subcarriers used. A simple segmentation method is to set different numbers of segments based on the number of effective subcarriers. For example, when the number of effective subcarriers is between 400 and 500, it is divided into five segments. For the HPLC frequency band 0, which contains 411 effective data subcarriers used in this embodiment, the effective data subcarrier indices 0-79 can be set as the first segment, 80-159 as the second segment, 160-239 as the third segment, 240-319 as the fourth segment, and 320-410 as the fifth segment. Calculate the average SNR for each of the five different frequency bands. For example, in this embodiment, the average SNRs for the five frequency segments are calculated as -2 dB, 6 dB, 8 dB, 12 dB, and 8 dB, respectively.
[0076] Step 4: Using the root mean square delay spread and average SNR calculated in Steps 2 and 3, determine the filtering configuration for Channel Estimation Method 1 based on the lookup table obtained from large-scale simulation. For example, when the channel estimation method is LMMSE, a possible lookup table is shown in Tables 1 and 2 below. Table 1 is the lookup table for root mean square delay spread and filtering interval for Channel Estimation Method 1, and Table 2 is the lookup table for signal-to-noise ratio and filtering order for Channel Estimation Method 1 (LMMSE). In this embodiment, a 16th-order filtering interval is selected for frequency band 1. The filtering method uses a 12th-order filter interval for frequency bands 2, 3, and 5. The filtering method, for frequency band 4, selects an 8th-order filter interval of 1. The filtering method.
[0077] Table 1
[0078] Table 2
[0079] Step 5: Based on the LMMSE filtering configuration determined in Step 4, generate Wiener filter coefficients or select pre-stored coefficients. The specific coefficient generation method is beyond the scope of this document. Following the configuration selected in Step 4, perform LMMSE filtering on every other subcarrier of the 411 data subcarriers in the State Grid HPLC band 0 (2-12 MHz).
[0080] Step 6: Perform linear interpolation estimation on the remaining half of the 411 data subcarriers in the State Grid HPLC band 0 (2 - 12 MHz).
[0081] In some embodiments, the present invention provides a power line orthogonal frequency division multiplexing (OFDM) communication receiving device, comprising: one or more processors; a memory; and one or more computer programs; wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and the one or more computer programs include instructions for performing the methods described above.
[0082] Verification Example To verify the effectiveness of this embodiment, the following simulation experiment can be performed: Simulation environment: A power line channel simulation platform conforming to the IEEE 1901 standard was built to simulate various typical power line channel environments, including additive white Gaussian noise (AWGN) channels and multipath fading channels.
[0083] Comparison scheme: The adaptive filtering scheme of this embodiment is compared with the traditional all-subcarrier fixed-order LMMSE filtering scheme.
[0084] Performance metrics: The main focus is on the frame error rate (FER) and the computational complexity of the channel estimation module (measured by the number of complex multiplications required).
[0085] Expected Results: Under similar FER performance conditions, record and compare the computational complexity of the two schemes. The expected validation framework is as follows: Figure 5 , Figure 6 As shown, the frame error rate of the filtering method in this embodiment is significantly lower than that of the traditional full subcarrier fixed-order LMMSE filtering scheme in both AWGN and multipath channels. Figure 5 and Figure 6 The physical blocks used are all 72, the number of copies is 1, the modulation method is QPSK, and the code rate is 1 / 2.
[0086] Optionally, a channel quality assessment step can be added before or after steps S2 and S3. For example, if the calculated noise power σ_n^2 exceeds an extremely high threshold, or the initial channel response is abnormal (such as a sharp change in amplitude), it can be determined that the current channel conditions are extremely poor. In this case, the adaptive process can be ignored, and a preset, most robust fixed filtering configuration (such as the minimum application interval and the highest filtering order) can be directly used for channel estimation to ensure the most basic communication performance.
[0087] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, several equivalent substitutions or obvious modifications can be made without departing from the concept of the present invention, and all such modifications, achieving the same performance or purpose, should be considered within the scope of protection of the present invention.
Claims
1. An adaptive channel estimation method for power line OFDM communication systems, characterized in that, Includes the following steps: S1. Obtain the initial frequency domain response estimate of the current channel; S2. Calculate the first channel parameter to characterize the channel frequency selective fading intensity based on the initial frequency domain response estimate; S3. Calculate the second channel parameter used to characterize the channel noise level based on the initial frequency domain response estimate; S4. Based on the first channel parameters and the second channel parameters, dynamically determine the filtering configuration of the first type of channel estimation method. The filtering configuration includes the application interval and filtering order of the first type of channel estimation method. The configuration lookup table records the filtering configuration that minimizes the total computational complexity under different channel states while meeting the preset channel estimation performance threshold. S5. Using the subcarriers determined by the application interval as nodes, a first type of channel estimation method with first filtering capability is used and the filtering order is used to perform channel estimation on the corresponding node subcarriers to obtain the node subcarrier channel estimation value. S6. For non-node subcarriers, a second type of channel estimation method with lower computational complexity than the first type of channel estimation method is used, and channel estimation is performed based on the channel estimation values of the adjacent node subcarriers.
2. The method according to claim 1, characterized in that, Step S1 specifically includes the following steps: S11. Receive the time-domain signal of the preamble sequence PREAM or training sequence TF in the power line communication system. S12. Perform a fast Fourier transform on the time-domain signal to obtain the frequency-domain received data; S13. Divide the frequency domain received data by the known frequency domain transmitted data to obtain the least squares estimate on each subcarrier, which is used as the initial frequency domain response estimate.
3. The method according to claim 2, characterized in that, In step S12, before performing a fast Fourier transform on the time-domain signal, the following steps are also included: If the PREAM or TF is repeatedly transmitted by multiple identical OFDM symbols, the time-domain signals of the multiple symbols are first averaged in the time domain, and then a fast Fourier transform is performed.
4. The method according to claim 1, characterized in that, In step S2, calculating the first channel parameter used to characterize the channel frequency-selective fading intensity specifically includes the following steps: S21. Perform a fast inverse Fourier transform on the initial frequency domain response estimate to obtain the time-delay domain channel impulse response; S22. Calculate the power delay spectrum based on the time-delay domain channel impulse response; S23. Identify and filter out effective paths from the power delay spectrum to obtain a power delay spectrum containing only effective paths; S24. Calculate the root mean square delay spread based on the power delay spectrum containing only the effective path, and use it as the first channel parameter.
5. The method according to claim 4, characterized in that, The power line OFDM communication system is a multiple-input multiple-output system, and step S22 specifically includes the following steps: Calculate the power delay spectrum between each transmit and receive antenna pair; In step S23, identifying and filtering effective paths from the power delay spectrum specifically includes the following steps: The power delay spectrum of each transmit and receive antenna pair is averaged to obtain the average power delay spectrum, and the effective path is identified from the average power delay spectrum.
6. The method according to claim 1, characterized in that, Step S2 specifically includes the following steps: Based on the initial frequency domain response estimate, the frequency domain amplitude crossover rate is calculated and used as the first channel parameter.
7. The method according to claim 1, characterized in that, Step S3, calculating the second channel parameter used to characterize the channel noise level, specifically includes the following steps: S31. Divide the channel estimation bandwidth into multiple sub-bandwidths; S32. Calculate the estimated signal-to-noise ratio within each sub-bandwidth and use it as the second channel parameter corresponding to each sub-bandwidth. Accordingly, the query of the preset configuration lookup table and the subsequent channel estimation steps are executed independently for each sub-bandwidth.
8. The method according to claim 1, characterized in that, The first type of channel estimation method is linear minimum mean square error filtering or transform domain filtering; the second type of channel estimation method is linear interpolation, cubic spline interpolation or Gaussian interpolation.
9. A power line orthogonal frequency division multiplexing (OFDM) communication receiving device, characterized in that, include: One or more processors; Memory; One or more computer programs; wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, the one or more computer programs including instructions for performing the method as claimed in any one of claims 1 to 8.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 8.