Cognitive radio based multicarrier wideband anti-jamming communication system

Abstract

The present application relates to the technical field of wireless communication, in particular to a multi-carrier broadband anti-interference communication system based on cognitive radio, which comprises a spectrum sensing module, an interference identification module, a subcarrier management module, a parameter configuration module and a signal generation module. The spectrum sensing module obtains the received signal strength and noise floor of each subcarrier in the target frequency band. The interference identification module processes the spectrum data by calling an improved hidden Markov model, which constructs a state transition constraint matrix based on the spectrum correlation between adjacent subcarriers, generates an interference state label sequence for each subcarrier, and improves the interference identification accuracy. The subcarrier management module screens out idle subcarriers to form an available set. The parameter configuration module adaptively configures the modulation order and coding rate for each idle subcarrier, realizing fine resource adaptation. The signal generation module finally generates a multi-carrier modulation transmission signal.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a multi-carrier broadband anti-interference communication system based on cognitive radio. Background Technology

[0002] In broadband wireless communication systems employing multi-carrier technologies such as orthogonal frequency division multiplexing (OFDM), the entire operating frequency band is divided into multiple parallel subcarriers. Broadband interference signals in complex electromagnetic environments may occupy only a portion of this frequency band, exhibiting time-varying and frequency-varying characteristics. Traditional broadband anti-interference communication systems often employ spectrum sensing technology to detect available spectrum gaps when dealing with such interference. A common approach is for the system to independently detect the energy of each subcarrier, comparing the received signal strength with a fixed or adaptive threshold to determine whether the subcarrier is occupied by interference. However, due to the randomness of wireless channel noise, this instantaneous, independent decision-making method is prone to misjudgments due to noise fluctuations, misclassifying an interfered subcarrier as idle, leading to transmission errors, or misclassifying strong noise pulses as interference, resulting in wasted spectrum resources. The reliability of the decision decreases, especially under low signal-to-interference-plus-noise ratio (SNR) conditions. Another approach is to introduce statistical processing in the time dimension, such as observing a single subcarrier for a longer period and making decisions based on statistical quantities, but this sacrifices the system's response speed to changes in interference states.

[0003] After identifying idle subcarriers, the system needs to allocate resources and configure parameters. Conventional methods tend to treat all available idle subcarriers as a single resource block, employing a uniform and relatively conservative modulation and coding scheme to adapt to the worst subchannel conditions. This approach fails to fully utilize the differences in channel quality between different subcarriers, limiting the system's throughput. Another improved approach is to allocate appropriate power or select different modulation schemes for different subcarriers based on the historical average or instantaneous measurement of the subcarrier's signal-to-noise ratio. However, its parameter configuration strategy is usually based on an ideal Gaussian channel model or a simple binary treatment of interference. When interference is correlated in the frequency domain, manifesting as continuous or regular occurrences on adjacent subcarriers, existing methods fail to effectively utilize this frequency domain structure information to improve the accuracy of interference state identification. Furthermore, in dynamic interference environments, parameter configuration based solely on the current instantaneous channel measurement may fail due to changes in interference at the next moment, leading to drastic fluctuations in link performance. Therefore, how to more accurately characterize and identify correlated interference state patterns in the broadband spectrum, and how to achieve more refined and robust dynamic adaptation of transmission parameters on this basis, are key issues for improving the communication performance of multi-carrier broadband systems in complex electromagnetic environments. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of existing technologies and propose a multi-carrier broadband anti-interference communication system based on cognitive radio.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a multi-carrier broadband anti-interference communication system based on cognitive radio, comprising: The spectrum sensing module acquires a set of broadband spectrum sampling data of the target communication frequency band within the current time window. The set of broadband spectrum sampling data includes the received signal strength indication value and noise floor estimate value corresponding to each of the multiple subcarriers. The interference identification module calls an improved hidden Markov model to perform interference state sequence decoding on the broadband spectrum sampling data set, generating an interference state label sequence for each subcarrier within the current time window. The improved hidden Markov model constructs a state transition constraint matrix based on the spectral correlation between adjacent subcarriers. The subcarrier management module selects subcarriers with idle interference status labels from the plurality of subcarriers based on the interference status label sequence, thereby forming a set of available subcarriers; The parameter configuration module performs adaptive modulation and coding parameter configuration processing on each subcarrier in the available subcarrier set to generate the modulation order and coding rate corresponding to each subcarrier. The signal generation module allocates the data stream to be transmitted to each subcarrier in the set of available subcarriers according to the modulation order and coding rate, thereby generating a multi-carrier modulated transmission signal.

[0006] As a further aspect of the present invention, the step of calling an improved Hidden Markov Model to perform interference state sequence decoding on the broadband spectrum sampling data set to generate an interference state label sequence for each subcarrier within the current time window includes: The received signal strength indication values ​​of each subcarrier at multiple sampling times within the current time window are extracted from the broadband spectrum sampling data set to form the observation time series of the subcarrier; The noise basis estimate of each subcarrier at multiple sampling times within the current time window is extracted from the broadband spectrum sampling data set to form the noise basis time series of the subcarrier; For each subcarrier, the ratio of the received signal strength indication value at each sampling time in the observation time series to the noise basis estimate value at the corresponding sampling time in the noise basis time series is calculated to generate the signal-to-noise ratio time series of the subcarrier; The signal-to-noise ratio time series of each subcarrier is input into the improved hidden Markov model as the observation sequence; The improved Hidden Markov Model performs Viterbi decoding on the observation sequence based on a preset state transition probability matrix, observation probability matrix, and initial state probability distribution. The off-diagonal elements in the state transition probability matrix are adaptively adjusted according to the correlation between the interference state labels of adjacent subcarriers. The Viterbi decoding operation outputs the interference state label that maximizes the joint probability at each sampling time. The interference state labels at all sampling times are arranged in chronological order to generate the interference state label sequence, wherein each interference state label is either occupied or idle.

[0007] As a further aspect of the present invention, the improved hidden Markov model constructs a state transition constraint matrix based on the spectral correlation between adjacent subcarriers, including: Obtain a subcarrier index sequence formed by arranging all subcarriers within the target communication frequency band in ascending order of frequency; For each subcarrier index in the subcarrier index sequence, determine the preceding and following subcarrier indices adjacent to the subcarrier index to form a set of adjacent subcarrier pairs; For each adjacent subcarrier pair, the co-occurrence frequency of two subcarriers in the adjacent subcarrier pair being occupied by the interference source at the same time, and the co-occurrence frequency of two subcarriers being in an idle state at the same time are counted from the historical spectrum monitoring data. The state consistency coefficient of adjacent subcarrier pairs is calculated based on the co-occurrence frequency. The state consistency coefficient is used to characterize the probability that the interference state labels of adjacent subcarrier pairs are the same. The state consistency coefficient is used as a weighting factor and multiplied by the corresponding element in the preset basic state transition probability matrix to generate the value of each element in the state transition constraint matrix. Normalize each row vector in the state transition constraint matrix so that the sum of all elements in each row vector equals the value one.

[0008] As a further aspect of the present invention, the adaptive adjustment process of the improved hidden Markov model includes: At the start of the current time window, the state transition probability matrix is ​​initialized to the state transition constraint matrix; After processing each sampling moment within the current time window, obtain the interference state labels obtained from decoding all subcarriers at the sampling moment to form a full-band state vector; The full-band state vector is scanned by a sliding window according to the subcarrier index order, and the label consistency ratio of adjacent subcarrier pairs is counted within the sliding window. The label consistency ratio is compared with a preset consistency threshold. When the label consistency ratio is lower than the consistency threshold, the values ​​of the diagonal elements in the state transition constraint matrix are reduced, while the values ​​of the off-diagonal elements in the state transition constraint matrix are increased. The adjusted state transition constraint matrix is ​​then normalized again to generate an updated state transition probability matrix, which is used for the Viterbi decoding operation at the next sampling time.

[0009] As a further aspect of the present invention, adaptive modulation and coding parameter configuration processing is performed on each subcarrier in the available subcarrier set to generate the modulation order and coding rate corresponding to each subcarrier, including: Select a target subcarrier from the available subcarrier set, and extract the end value of the signal-to-noise ratio time series of the target subcarrier at the last sampling time in the current time window, as the current signal-to-noise ratio estimate of the target subcarrier; Obtain a mapping table between multiple preset signal-to-noise ratio ranges and modulation and coding schemes. Each modulation and coding scheme contains a modulation order value and a coding rate value. Traverse all signal-to-noise ratio (SNR) intervals in the mapping table, and determine the modulation and coding scheme corresponding to the SNR interval to which the current SNR estimate belongs as the candidate modulation and coding scheme for the target subcarrier; The interference status labels of the target subcarrier at all sampling times within the current time window are extracted from the interference status label sequence. The proportion of the sampling times with the interference status label being in the occupied state to the total number of sampling times in the current time window is counted as the interference duty cycle of the target subcarrier. The coding rate in the candidate modulation and coding scheme is adjusted downward based on the interference duty cycle to generate a corrected coding rate value, wherein the magnitude of the downward adjustment is positively correlated with the magnitude of the interference duty cycle. The modulation order of the target subcarrier is determined as the modulation order value in the candidate modulation and coding scheme, the coding rate of the target subcarrier is determined as the corrected coding rate value, and the operation is repeated for the next subcarrier in the set of available subcarriers until all subcarriers obtain the corresponding modulation order and coding rate.

[0010] As a further aspect of the present invention, the data stream to be transmitted is allocated to each subcarrier in the set of available subcarriers according to the modulation order and coding rate to generate a multi-carrier modulated transmission signal, including: Obtain the original bit sequence of the data stream to be sent, and divide the original bit sequence into multiple data frames according to a preset frame structure. Each data frame contains a frame header synchronization sequence and a frame body data sequence. For each subcarrier in the set of available subcarriers, channel coding is performed on the frame data sequence to be carried by the subcarrier according to the coding code rate corresponding to the subcarrier to generate a coded bit sequence; A constellation mapping operation is performed on the encoded bit sequence according to the modulation order corresponding to the subcarrier to generate a complex symbol sequence corresponding to the subcarrier; Arrange the complex symbols of all subcarriers within the same symbol period in subcarrier index order to form an orthogonal frequency division multiplexing symbol vector; Perform an inverse fast Fourier transform operation on the orthogonal frequency division multiplexing symbol vector to generate a time-domain orthogonal frequency division multiplexing baseband signal; The time-domain orthogonal frequency division multiplexing baseband signal is sequentially subjected to parallel-to-serial conversion, digital-to-analog conversion, and radio frequency up-conversion to generate the multi-carrier modulated transmission signal.

[0011] As a further aspect of the present invention, an inverse fast Fourier transform operation is performed on the orthogonal frequency division multiplexing symbol vector to generate a time-domain orthogonal frequency division multiplexing baseband signal, specifically including: Extract the complex symbol sequence of all subcarriers in the current symbol period from the orthogonal frequency division multiplexing symbol vector; Calculate the length of the complex symbol sequence and compare it with the preset number of points in the Fast Fourier Transform; If the length of the complex symbol sequence is equal to the preset number of Fast Fourier Transform points, then the complex symbol sequence is directly used as the input sequence. If the length of the complex symbol sequence is less than the preset number of Fast Fourier Transform points, then a corresponding number of zero-value symbols are added to the high-frequency positions of the complex symbol sequence to make the length of the input sequence reach the preset number of Fast Fourier Transform points. The input sequence is input into the inverse fast Fourier transform calculation module in subcarrier index order; The inverse fast Fourier transform calculation module performs an inverse discrete Fourier transform operation on the input sequence to generate an initial baseband signal sequence in the time domain; A cyclic prefix insertion operation is performed on the initial baseband signal sequence in the time domain, truncating a specified length of sample points from the end of the sequence and copying them to the beginning of the sequence; After inserting the cyclic prefix, a windowing filter is applied to the resulting extended sequence to suppress out-of-band spectral leakage and generate the final time-domain orthogonal frequency division multiplexing baseband signal.

[0012] As a further aspect of the present invention, before performing constellation mapping operation on the encoded bit sequence according to the modulation order corresponding to the subcarrier to generate the complex symbol sequence corresponding to the subcarrier, the method further includes performing: Extract the interference status label of each subcarrier at each sampling moment within the current time window from the interference status label sequence, and determine the interference idle time distribution of each subcarrier in the time dimension; Based on the distribution of the interference idle time periods, the encoded bit sequence corresponding to each subcarrier is divided into multiple sub-data blocks, and the length of each sub-data block is matched with the number of bits that the subcarrier can transmit in the corresponding idle time period; In the process of constructing the orthogonal frequency division multiplexing symbol vector, only the complex symbols in the complex symbol sequence that are located within the distribution of the interference idle time period are filled into the corresponding symbol period positions; For the symbol period position in the orthogonal frequency division multiplexing symbol vector corresponding to the interference-occupied period, a preset empty symbol value is filled in so that the subcarriers during the interference-occupied period do not carry valid data.

[0013] As a further aspect of the present invention, after generating the multi-carrier modulated transmission signal, the method further includes performing: The receiving end processes the multi-carrier modulated transmitted signal to obtain the received broadband spectrum data set; Perform serial-to-parallel conversion and fast Fourier transform operations on the received broadband spectrum data set to extract the received complex symbols of each subcarrier in each symbol period; For each subcarrier, a soft demodulation operation is performed on the received complex symbols according to the modulation order corresponding to the subcarrier to generate a soft bit metric sequence for the subcarrier; The soft bit metric sequence is subjected to channel decoding based on the coding rate corresponding to the subcarrier to generate the decoded bit sequence of the subcarrier; The decoded bit sequences of all subcarriers are concatenated according to the data allocation order at the transmitting end to recover the original data stream to be transmitted.

[0014] As a further aspect of the present invention, after performing serial-to-parallel conversion and fast Fourier transform operations on the received broadband spectrum data set to extract the received complex symbols of each subcarrier in each symbol period, the method further includes performing: Extract the pilot symbol sequence located at the preset pilot subcarrier position from the received complex symbols; The pilot symbol sequence is multiplied symbol by symbol by symbol by the locally stored reference pilot symbol sequence to obtain the channel frequency response estimate at each pilot symbol position; Perform frequency domain interpolation on the channel frequency response estimate to generate channel frequency response estimates for all subcarriers in each symbol period; The channel frequency response estimate is used to perform frequency domain equalization on the received complex symbols of each subcarrier in each symbol period to generate equalized received complex symbols, which are used to replace the original received complex symbols in the soft demodulation operation.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: The improved Hidden Markov Model (HMM) constructs a state transition constraint matrix based on the spectral correlation between adjacent subcarriers, replacing the assumption in the traditional model that the state transition probability is consistent across all subcarriers. Spectral correlation arises because real-world broadband interference signals typically cover continuous or specific frequency ranges, leading to adjacent subcarriers tending to have similar or strongly correlated interference states. Encoding this prior knowledge into the state transition constraint matrix allows the model to prioritize state transition paths conforming to this spectral correlation during sequence decoding. This approach expands the decision-making basis for interference identification from isolated, instantaneous subcarrier energy detection values ​​to the entire sequence containing frequency domain structure information. During decoding, even if the received signal strength indicator value of a subcarrier approaches the decision threshold due to instantaneous noise, the model will make a more frequency-domain-compliant inference about the interference state at that point based on the states of its adjacent subcarriers. This effectively suppresses isolated false alarms caused by random fluctuations in noise, making the interference state label sequence generated by decoding more reflective of the true distribution contour of interference in the frequency domain, reducing the probability of false alarms and missed alarms in the interference state identification results, and improving the robustness of interference identification, especially interference with continuous frequency domain characteristics, under low signal-to-interference-plus-noise ratio conditions.

[0016] Based on the subcarrier interference status label sequence generated from the real-time interference identification results, adaptive modulation and coding parameter configuration is performed on each of the selected idle subcarriers. The parameter configuration not only considers the current received signal strength and noise floor of the subcarrier, but its core lies in combining the implicit information of the subcarrier's historical interference status and future trends. Through the status sequence output by the interference identification module, the system can distinguish between subcarriers with good channel quality but potentially about to be overwhelmed by interference, and subcarriers that have experienced interference but have become stable. For subcarriers assessed as stable and idle with excellent channel conditions, an aggressive transmission mode with high-order modulation and high code rate can be adopted to maximize their data transmission capacity. For idle subcarriers with average channel quality or questionable stability, a lower-order modulation and coding scheme with stronger error correction capabilities is used to ensure transmission reliability. This independent and refined parameter adaptation mechanism for each idle subcarrier abandons the approach of using a uniform and conservative configuration for all idle subcarriers. It fully utilizes the potentially significant differences in channel quality between fragmented spectrum pieces caused by interference, achieving a refined improvement in spectral efficiency. Meanwhile, since the parameter configuration closely follows the interference state prediction based on sequence decoding, when the interference migrates or changes in its form in the frequency domain, the system can proactively adjust the transmission parameters of the relevant subcarriers, so that the data transmission process can still maintain high throughput and stable link performance in the environment of dynamic interference changes, thereby enhancing the overall adaptability and robustness of the communication system. Attached Figure Description

[0017] Figure 1 This is a timing diagram of the multi-carrier broadband anti-interference communication system based on cognitive radio described in this invention; Figure 2 A flowchart for constructing the state transition constraint matrix of an improved Hidden Markov Model; Figure 3 A flowchart for generating multi-carrier modulated transmission signals. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0019] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0020] See Figure 1 An implementation of a multi-carrier broadband anti-interference communication system based on cognitive radio includes: a spectrum sensing module, an interference identification module, a subcarrier management module, a parameter configuration module, and a signal generation module. The spectrum sensing module continuously monitors the target communication frequency band, collecting the received signal strength indication values ​​and background noise floor estimates of all subcarriers within the band, periodically, using the current time window as the cycle, forming a broadband spectrum sampling data set containing multiple subcarriers and multiple sampling times. The interference identification module receives this broadband spectrum sampling data set and internally runs an improved Hidden Markov Model. This model utilizes the correlation between adjacent subcarriers in the spectrum to construct a state transition constraint matrix. Through processing the input data, it decodes and generates an interference state label sequence for each subcarrier in the time dimension within the current time window. Each label identifies whether the subcarrier is in an occupied or idle state at the corresponding sampling time.

[0021] The subcarrier management module, based on the generated interference status label sequence, filters out all subcarriers marked as idle from all subcarriers, forming a set of available subcarriers. This set represents the spectrum resources that can be safely used for communication at the current moment. The parameter configuration module, for each member in the available subcarrier set, independently calculates and allocates a suitable modulation scheme and channel coding strength based on its specific channel conditions and interference history; that is, it determines a modulation order and a coding rate for each available subcarrier. The signal generation module is ultimately responsible for executing the communication task. It divides and maps the data stream to be transmitted onto each subcarrier in the available subcarrier set according to the modulation order and coding rate configured for each subcarrier, and synthesizes it into the final transmission signal through multi-carrier modulation technology, thereby achieving reliable data transmission in a broadband environment with interference.

[0022] In one embodiment of the present invention, the broadband spectrum sampling data set acquired by the spectrum sensing module is transmitted to the interference identification module. The interference identification module extracts the received signal strength indication value of each subcarrier at each sampling moment within the current time window from the set, arranges these values ​​in time sequence to form an independent observation time series for each subcarrier, and extracts the noise basis estimate value of the corresponding subcarrier at the same moment to form its noise basis time series. For each subcarrier, the ratio of the received signal strength indication value at each moment in its observation time series to the noise basis estimate value at the same moment in the noise basis time series is calculated, thereby generating a signal-to-noise ratio (SNR) time series reflecting the channel quality of the subcarrier over time. The SNR time series of each subcarrier is used as an observation sequence and input into an improved hidden Markov model. This model has a predefined set of states, including occupied and idle states, and presets an initial state probability distribution and an observation probability matrix. The core of the model lies in its state transition probability matrix, whose off-diagonal elements are dynamically and adaptively adjusted according to the correlation between the interference state labels of adjacent subcarriers. The model uses the Viterbi decoding algorithm to process the input observation sequence. Based on the current state transition probability matrix, observation probability matrix, and initial state probability distribution, this algorithm calculates and finds a hidden state path that maximizes the probability of occurrence of the entire observation sequence. The output of the Viterbi decoding operation is the interference state label that maximizes the joint probability at each sampling time. Arranging these labels in the order of sampling time constitutes the interference state label sequence for the subcarrier within the current time window. Each label in the sequence represents either an occupied state or an idle state.

[0023] In practical implementation, the interference identification module uses an improved Hidden Markov Model to decode the interference state sequence of the broadband spectrum sampling data set, generating an interference state label sequence for each subcarrier within the current time window. Specifically, the interference identification module extracts the received signal strength indication (RSI) values ​​for each subcarrier at multiple sampling times within the current time window from the broadband spectrum sampling data set. These values ​​are arranged chronologically to form the observation time series of the subcarrier. Simultaneously, the interference identification module extracts the noise floor estimate values ​​for each subcarrier at the same multiple sampling times from the broadband spectrum sampling data set, forming the noise floor time series of the subcarrier. For each subcarrier, the interference identification module calculates the ratio of the RSI value at each sampling time in the observation time series to the noise floor estimate value at the corresponding sampling time in the noise floor time series. This calculation generates the signal-to-noise ratio (SNR) time series of the subcarrier. In some embodiments, the target communication frequency band is 1.9 GHz to 2.0 GHz, which is uniformly divided into 100 subcarriers, each with a bandwidth of 10 kHz. The current time window length is 10 milliseconds, and sampling is performed at 1-millisecond intervals. The broadband spectrum sampling data set collected by the spectrum sensing module at 10 sampling times includes the received signal strength indication value sequence and noise floor estimation value sequence for each of the 100 subcarriers. The interference identification module calculates a signal-to-noise ratio time series of length 10 for each subcarrier.

[0024] The interference identification module inputs the signal-to-noise ratio (SNR) time series of each subcarrier into the improved Hidden Markov Model (HMM) as the observation sequence. The improved HMM performs Viterbi decoding on the observation sequence based on a preset state transition probability matrix, observation probability matrix, and initial state probability distribution. The off-diagonal elements in the state transition probability matrix are adaptively adjusted according to the correlation between interference state labels of adjacent subcarriers. The observation probability matrix of the improved HMM defines the probability distribution of observing a specific SNR value under a given interference state label. In a specific implementation, for an interference state label s, its observation probability density function can be modeled as a Gaussian distribution, expressed as:

[0025] in: This represents the signal-to-noise ratio value observed at sampling time t. This represents the observed mean signal-to-noise ratio under the interference state label s. The standard deviation of the signal-to-noise ratio (SNR) observations under the interference state label s is represented by the formula used to calculate the transmission probability during the Viterbi decoding process. The Viterbi decoding operation outputs the interference state label that maximizes the joint probability at each sampling time. The improved Hidden Markov Model (HMM) arranges the interference state labels at all sampling times in chronological order to generate an interference state label sequence, where each interference state label takes the value of either occupied or idle. In some embodiments, for subcarrier index 50, its input SNR time series is [15.2, 14.8, 3.1, 2.9, 2.8, 13.9, 14.5, 15.0, 14.7, 15.1] (unit: dB). After the Viterbi decoding operation of the improved HMM, the output interference state label sequence is [idle, idle, occupied, occupied, occupied, idle, idle, idle, idle, idle, idle]. This indicates that within this time window, subcarrier 50 is determined to be interfered with in the middle three sampling times.

[0026] Optionally, during the initialization phase of the Viterbi decoding operation, the improved Hidden Markov Model sets initial probabilities for each possible state, and the initial state probability distribution can be set to a uniform distribution. In a specific implementation, the state transition probability matrix is ​​a 2x2 matrix, with its rows and columns corresponding to the idle and occupied states, respectively. Each element of the matrix... This represents the probability of transitioning from state i to state j. The generation of the observation probability matrix relies on statistical analysis of historical training data to obtain the statistical characteristics of the signal-to-noise ratio observations under different interference state labels. The Viterbi decoding operation recursively calculates the maximum probability path and its probability value for each state at each time step, and after processing the entire observation sequence, backtracks to obtain the globally optimal interference state label sequence.

[0027] It is understandable that the Viterbi decoding operation searches for the optimal path by maintaining a path metric and a backtracking pointer matrix. For an observation sequence of length T, the improved Hidden Markov Model provides the path metric for state j at time t. Through recursive formula The calculation is performed, where i iterates through all possible previous time-states. The interference identification module ultimately generates an independent interference state label sequence for each subcarrier within the current time window. Optionally, after generating the interference state label sequences for all subcarriers, the interference identification module can output these sequences as a two-dimensional array, where the row indices correspond to the subcarrier indices, the column indices correspond to the sampling time indices, and each element in the array is an interference state label. It can be understood that the generation of the interference state label sequence is the foundation for subsequent subcarrier management and parameter configuration, and its accuracy directly depends on the improved Hidden Markov Model's ability to model and decode the time-varying characteristics of the spectrum.

[0028] In one embodiment of the present invention, the construction of the state transition constraint matrix in the improved hidden Markov model depends on the analysis of the spectral correlation of adjacent subcarriers, see reference. Figure 2 The system acquires a subcarrier index sequence formed by arranging all subcarriers within the target communication frequency band in ascending order of frequency. For each subcarrier index in this sequence, the preceding and following indices in frequency are determined, thus forming a series of adjacent subcarrier pairs. Based on historical spectrum monitoring data, the co-occurrence frequencies of two subcarriers being simultaneously labeled as occupied by the same interference source, and the co-occurrence frequencies of two subcarriers simultaneously being in an idle state, are statistically analyzed for each adjacent subcarrier pair. Based on these statistically analyzed co-occurrence frequencies, a state consistency coefficient is calculated, which quantifies the probability that the pair of adjacent subcarriers will have the same interference state label in history. This state consistency coefficient is used as a weighting factor and multiplied by the corresponding matrix element in a preset basic state transition probability matrix to generate the initial element values ​​of the state transition constraint matrix. To satisfy the normalization of the probability matrix, each row vector in the state transition constraint matrix is ​​normalized to ensure that the sum of all elements in each row is one. During model operation, the state transition probability matrix is ​​adaptively adjusted. At the beginning of the current time window, the state transition probability matrix is ​​initialized to the state transition constraint matrix constructed above. After completing Viterbi decoding at each sampling time, the system collects the interference state labels obtained from decoding all subcarriers at that time, forming a full-band state vector reflecting the state across the entire frequency band. Following the subcarrier index order, a sliding window is used to scan this vector, and the proportion of interference state labels matching between adjacent subcarrier pairs is counted within the window. This label matching proportion is compared to a preset matching threshold. If the proportion is lower than the threshold, the values ​​of the diagonal elements in the state transition constraint matrix are decreased, while the values ​​of the off-diagonal elements are increased, indicating that the model considers the probability of a state transition to be increased. The adjusted state transition constraint matrix is ​​then subjected to row vector normalization again to generate an updated state transition probability matrix, which is used for the Viterbi decoding operation at the next sampling time.

[0029] In specific implementations, the improved Hidden Markov Model (HMM) constructs a state transition constraint matrix based on the spectral correlation between adjacent subcarriers. The improved HMM obtains a subcarrier index sequence formed by arranging all subcarriers within the target communication band in ascending order of frequency. In specific implementations, the subcarrier index sequence is a continuous integer array from 0 to N-1, where N represents the total number of subcarriers, and the order of the index values ​​directly corresponds to the ascending order of the subcarrier frequencies. For each subcarrier index in the subcarrier index sequence, the improved HMM determines the preceding and following subcarrier indices adjacent to that index, forming a set of adjacent subcarrier pairs. In some embodiments, the total number of subcarriers N is 128. For a subcarrier with index k, its adjacent subcarrier pairs include (k, k-1) and (k, k+1). When k is 0 or N-1, the adjacent subcarrier pair contains only one valid pair, and the set of adjacent subcarrier pairs contains a total of 255 adjacent subcarrier pairs formed by all subcarriers.

[0030] For each adjacent subcarrier pair, the improved Hidden Markov Model (HMM) statistically analyzes historical spectrum monitoring data to determine the co-occurrence frequencies of both subcarriers being simultaneously occupied by an interference source and the co-occurrence frequencies of both subcarriers being simultaneously idle. The state consistency coefficient of the adjacent subcarrier pair is then calculated based on these co-occurrence frequencies. The state consistency coefficient characterizes the probability that adjacent subcarrier pairs have the same interference state label. In a specific implementation, the formula for calculating the state consistency coefficient can be expressed as:

[0031] in: This represents the state consistency coefficient between subcarrier m and subcarrier n, a pair of adjacent subcarriers. This represents the co-occurrence frequency of subcarrier m and subcarrier n being simultaneously marked as occupied, obtained from historical spectrum monitoring data. This represents the co-occurrence frequency of subcarrier m and subcarrier n being simultaneously marked as idle, obtained from historical spectrum monitoring data. H represents the total number of samples in the historical spectrum monitoring data. This can be understood as the state consistency coefficient. The value ranges from 0 to 1. A higher value indicates that subcarrier m and subcarrier n have a greater probability of having the same interference state label in history, reflecting a stronger correlation between the two in the frequency domain.

[0032] The improved Hidden Markov Model uses the state consistency coefficient as a weighting factor, multiplied by the corresponding element in the preset base state transition probability matrix, to generate the value of each element in the state transition constraint matrix. The preset base state transition probability matrix is ​​a 2x2 matrix, with its rows and columns corresponding to the idle state (S0) and occupied state (S1), respectively. The base state transition probability matrix is ​​typically set as a diagonal dominance matrix. In specific implementations, for state transitions involving subcarriers m and n, the elements in the state transition constraint matrix corresponding to the transition from state i to state j... It can be calculated Obtain, among which These are the corresponding elements in the basic state transition probability matrix. The improved Hidden Markov Model normalizes each row vector in the state transition constraint matrix so that the sum of all elements in each row vector equals the value one. Optionally, normalization is achieved by dividing each element in each row vector by the sum of all elements in that row. After normalization, it becomes .

[0033] The adaptive adjustment process of the improved Hidden Markov Model initializes the state transition probability matrix to the state transition constraint matrix at the beginning of the current time window. In some embodiments, the state consistency coefficient between subcarrier 1 and subcarrier 2 is calculated to be 0.8, and the basic state transition probability matrix is ​​[[0.9,0.1],[0.1,0.9]]. The elements of the unnormalized state transition constraint matrix are calculated according to the formula. , , , The normalized state transition probability matrix is ​​initialized to [[0.82,0.18],[0.18,0.82]]. After processing each sampling moment within the current time window, the improved Hidden Markov Model obtains the interference state labels decoded from all subcarriers at the sampling moment, forming a full-band state vector. A sliding window scan is performed on the full-band state vector according to the subcarrier index order, and the label consistency ratio of adjacent subcarrier pairs is counted within the sliding window. It can be understood that the size of the sliding window can be set to cover 5 consecutive subcarrier indices, and the label consistency ratio refers to the proportion of pairs of adjacent subcarrier pairs with the same interference state label out of the total number of pairs within the sliding window.

[0034] The improved Hidden Markov Model (HMM) compares the label consistency ratio with a preset consistency threshold. When the label consistency ratio is lower than the threshold, it decreases the values ​​of the diagonal elements in the state transition constraint matrix while increasing the values ​​of the off-diagonal elements. In a specific implementation, the consistency threshold can be set to 0.6. If the label consistency ratio calculated after a certain sampling time is 0.4, which is lower than the consistency threshold, the improved HMM temporarily adjusts the state transition constraint matrix. Optionally, the adjustment operation multiplies each diagonal element of the matrix by an attenuation factor β less than 1 (e.g., β=0.9) and adds a compensation value γ (e.g., γ=0.05) to each off-diagonal element. The improved HMM then performs row vector normalization on the adjusted state transition constraint matrix again to generate an updated state transition probability matrix, which is used for the Viterbi decoding operation at the next sampling time. This adaptive adjustment process enables the state transition probability matrix to dynamically reflect the real-time changes in the state consistency between adjacent subcarriers at the current time.

[0035] In one embodiment of the invention, the parameter configuration module performs adaptive modulation and coding parameter configuration on subcarriers in the available subcarrier set. For a target subcarrier in the set, the module extracts the value of the last sampling moment within the current time window from its signal-to-noise ratio (SNR) time series as the current SNR estimate for that subcarrier. The system internally stores a mapping table that defines multiple consecutive SNR intervals, each interval associated with a specific modulation and coding scheme. Each scheme contains a defined modulation order value and a coding rate value.

[0036] The module iterates through this mapping table, identifying the modulation and coding schemes corresponding to the SNR intervals into which the current SNR estimate falls as candidate modulation and coding schemes for the target subcarrier. From the interference state label sequence of this subcarrier, the module calculates the proportion of sampling times marked as occupied within the current time window to the total number of sampling times, obtaining the interference duty cycle of that subcarrier. This interference duty cycle is used to lower the coding rate in the candidate modulation and coding scheme; the magnitude of the correction is positively correlated with the size of the interference duty cycle—the more frequent the interference, the greater the reduction in coding rate to improve robustness. Finally, the modulation order of the target subcarrier is determined as the modulation order value in the candidate modulation and coding scheme, and its coding rate is the corrected coding rate value. The parameter configuration module repeats the above operations for each subcarrier in the available subcarrier set until all subcarriers have obtained independently configured modulation order and coding rate.

[0037] In specific implementation, the parameter configuration module performs adaptive modulation and coding parameter configuration processing on each subcarrier in the available subcarrier set. The parameter configuration module selects a target subcarrier from the available subcarrier set and extracts the end value of the signal-to-noise ratio (SNR) time series of the target subcarrier at the last sampling moment within the current time window. This end value is used as the current SNR estimate for the target subcarrier. In some embodiments, the available subcarrier set includes subcarriers with indices 2, 5, and 8. The parameter configuration module selects subcarrier 2 as the target subcarrier. The SNR time series of subcarrier 2 is [12.1, 11.8, 12.5, 11.9, 12.3, 12.0, 12.6, 12.2, 12.4, 12.1] (unit: dB). The parameter configuration module extracts the value of 12.1 dB from the last sampling moment, i.e., the 10th moment, as the current SNR estimate.

[0038] The parameter configuration module obtains a pre-defined mapping table between multiple signal-to-noise ratio (SNR) intervals and modulation / coding schemes. The mapping table defines multiple consecutive SNR intervals, each associated with a unique modulation / coding scheme. Each modulation / coding scheme contains a modulation order value and a coding rate value. See Table 1 for an example of a mapping table between SNR intervals and modulation / coding schemes.

[0039] Table 1: Mapping Table of Signal-to-Noise Ratio Range and Modulation / Coding Scheme

[0040] The parameter configuration module iterates through all signal-to-noise ratio (SNR) intervals in the mapping table and determines the modulation and coding scheme corresponding to the SNR interval to which the current SNR estimate belongs as a candidate modulation and coding scheme for the target subcarrier. In specific implementation, for a target subcarrier with a current SNR estimate of 12.1 dB, the parameter configuration module looks up Table 1 to determine that 12.1 dB falls within the interval [10.0, 15.0) dB. The parameter configuration module then determines the modulation scheme 16QAM and coding rate 2 / 3 corresponding to this interval as candidate modulation and coding schemes for the target subcarrier, where the modulation order is 4 (corresponding to 16QAM) and the candidate coding rate is 2 / 3.

[0041] The parameter configuration module extracts the interference state labels of the target subcarrier from the interference state label sequence for all sampling moments within the current time window. It then calculates the ratio of sampling moments with the interference state label "occupied" to the total number of sampling moments in the current time window; this ratio is used as the interference duty cycle of the target subcarrier. In some embodiments, the interference state label sequence of target subcarrier 2 is [idle, idle, occupied, idle, idle, idle, occupied, idle, idle, idle], the total number of sampling moments in the current time window is 10, and the number of sampling moments with the interference state label "occupied" is 2. The interference duty cycle of the target subcarrier is calculated as 2 / 10 = 0.2. Based on the interference duty cycle, the coding rate in the candidate modulation and coding scheme is adjusted downwards to generate a corrected coding rate value. In specific implementations, the magnitude of the downward adjustment is positively correlated with the size of the interference duty cycle. The corrected coding rate value is then calculated. It can be calculated using the formula:

[0042] in: This represents the corrected coding rate value. This represents the coding rate value among the candidate modulation and coding schemes retrieved from the mapping table. This indicates a preset correction factor used to control the intensity of the downsampling. This represents the calculated interference duty cycle of the target subcarrier. The parameter configuration module determines the modulation order of the target subcarrier as the modulation order value in the candidate modulation and coding scheme, and determines the coding rate of the target subcarrier as the corrected coding rate value. Optional, a preset correction coefficient. It can be set to 0.5 for candidate coding bitrate values. Furthermore, for the target subcarrier with an interference duty cycle of D=0.2, the corrected coding rate value .

[0043] The parameter configuration module repeats the operation for the next subcarrier in the available subcarrier set until all subcarriers have obtained the corresponding modulation order and coding rate. It can be understood that the parameter configuration module processes subcarrier 5 and subcarrier 8 in the available subcarrier set sequentially. For subcarrier 5, its current signal-to-noise ratio estimate is 18.5 dB, which falls within the interval [15.0, 20.0) dB according to lookup table 1. The candidate modulation and coding scheme is 64QAM (modulation order 6) and the coding rate is 3 / 4. The interference duty cycle of subcarrier 5 is 0.1, and the corrected coding rate is... It can be understood that the adaptive modulation and coding parameter configuration process assigns a suitable modulation and coding combination to each available subcarrier, and the reduction of the coding rate proactively addresses the historical interference statistical characteristics of the subcarrier. In specific implementations, the contents of the mapping table can be predefined or dynamically updated according to channel conditions and system requirements. Optionally, the interference duty cycle can be calculated based on a historical interference state label sequence over a longer time span to obtain more stable statistical characteristics. The parameter configuration module ultimately outputs a configuration list, where each item records the index of a subcarrier in the available subcarrier set, the modulation order configured for it, and the corrected coding rate value.

[0044] In one embodiment of the present invention, the signal generation module allocates the data stream to be transmitted to each subcarrier according to the configured parameters, see reference. Figure 3 The module acquires the original bit sequence of the data stream to be transmitted and divides it into multiple data frames according to a preset frame structure. Each data frame contains a frame header synchronization sequence and a frame body data sequence. For each subcarrier in the available subcarrier set, based on its configured coding rate, the module performs channel coding operations, such as convolutional coding or LDPC coding, on the frame body data sequence planned to be carried by that subcarrier, generating a coded bit sequence. To improve transmission reliability, the module utilizes information provided by the interference state label sequence before constellation mapping.

[0045] The module extracts the interference state label for each subcarrier at each sampling moment within the current time window from the sequence, thereby determining the interference idle period distribution of each subcarrier in the time dimension. Based on this distribution, the coded bit sequence corresponding to each subcarrier is divided into multiple sub-data blocks, the length of which precisely matches the number of bits that the subcarrier can transmit in a corresponding continuous idle period. According to the modulation order configured for the subcarrier, constellation mapping operations, such as QPSK or 16QAM, are performed on the segmented coded bit sequence (or sub-data blocks) to generate the complex symbol sequence corresponding to that subcarrier. When constructing the orthogonal frequency division multiplexing symbol vector, only those symbols in the complex symbol sequence that are located in the identified interference idle periods are filled into the corresponding symbol period positions. For the symbol period positions in the symbol vector corresponding to the interference occupied period, a preset empty symbol value is filled in, thereby ensuring that the subcarrier does not carry any valid data during the interference occupied period. In each symbol period, the complex symbols of all subcarriers in that period are arranged in order of their subcarrier indices to form an orthogonal frequency division multiplexing symbol vector. The symbol vector is subjected to an inverse fast Fourier transform (IFFT) to generate a time-domain signal. Specifically, the complex symbol sequence of all subcarriers in the current symbol period is extracted, and its length is compared with the preset number of IFFT points. If the length is insufficient, zeros are padded at high-frequency positions to achieve the required length, forming the input sequence. The input sequence is then fed into the IFFT calculation module to perform an inverse discrete Fourier transform (DFT), obtaining the initial baseband signal sequence in the time domain. A cyclic prefix insertion operation is performed on this initial sequence, and sample points of a specified length are truncated from the end of the sequence and copied to the beginning. A windowed filter operation is then performed on the extended sequence after the cyclic prefix insertion to suppress out-of-band spectral leakage, generating the final time-domain orthogonal frequency division multiplexing (OFDM) baseband signal. This baseband signal is then subjected to parallel-to-serial conversion, digital-to-analog conversion, and RF up-conversion processing to finally generate a multi-carrier modulated transmission signal.

[0046] In practice, the signal generation module allocates the data stream to be transmitted to each subcarrier in the available subcarrier set according to the modulation order and coding code rate. The signal generation module obtains the original bit sequence of the data stream to be transmitted and divides the original bit sequence into multiple data frames according to the preset frame structure. Each data frame contains a frame header synchronization sequence and a frame body data sequence. For each subcarrier in the available subcarrier set, the signal generation module performs channel coding operation on the frame body data sequence to be carried by the subcarrier according to the coding code rate corresponding to the subcarrier, and generates the coded bit sequence. In some embodiments, the available subcarrier set includes subcarriers with subcarrier indices 1, 3, and 4. The coding code rate configured for subcarrier 1 is 1 / 2 and the modulation order is 2 (QPSK). The coding code rate configured for subcarrier 3 is 2 / 3 and the modulation order is 4 (16QAM). The coding code rate configured for subcarrier 4 is 3 / 4 and the modulation order is 6 (64QAM). The data stream to be transmitted is divided into a sequence containing 10 data frames. The frame body data sequence of each data frame is 120 bits long. For subcarrier 1, its frame body data sequence to be carried is convolutionally encoded at a code rate of 1 / 2 to generate a encoded bit sequence of 240 bits in length.

[0047] Before performing constellation mapping on the encoded bit sequence based on the modulation order corresponding to the subcarrier, the signal generation module performs the following operations: The signal generation module extracts the interference state label for each subcarrier at each sampling moment within the current time window from the interference state label sequence, determines the interference idle time distribution for each subcarrier in the time dimension, and divides the encoded bit sequence corresponding to each subcarrier into multiple sub-data blocks based on the interference idle time distribution. The length of each sub-data block matches the number of bits that can be transmitted by the subcarrier in the corresponding idle time period. Refer to Table 2, which shows the interference idle time distribution and data block segmentation for a subcarrier.

[0048] Table 2: Distribution of Subcarrier Interference Idle Periods and Data Block Segmentation Table

[0049] In specific implementation, for subcarrier 1, its interference idle period distribution includes two consecutive periods: symbol period 1 to 3, and symbol period 6 to 9. The signal generation module calculates based on modulation order 2 (QPSK, 2 bits per symbol) that the number of bits that can be transmitted in the first idle period (3 symbol periods) is 3 symbol periods × 2 bits / symbol = 6 bits. However, considering that the actual system carries multiple OFDM symbols in one symbol period, Table 1 assumes that the number of symbol periods in each idle period corresponds to multiple OFDM symbols. The number of bits that can be transmitted needs to be calculated based on the actual symbol rate and modulation order. In the example, 144 bits can be transmitted in the first idle period. Therefore, the signal generation module divides the encoded bit sequence of subcarrier 1 into a first sub-data block of 144 bits for transmission in periods 1-3, and a second sub-data block of 192 bits for transmission in periods 6-9.

[0050] Based on the modulation order corresponding to the subcarrier, a constellation mapping operation is performed on the encoded bit sequence to generate a complex symbol sequence corresponding to the subcarrier. During the construction of the orthogonal frequency division multiplexing (OFDM) symbol vector, only complex symbols located within the interference idle period distribution of the complex symbol sequence are filled into the corresponding symbol period positions. For the symbol period positions in the OFDM symbol vector corresponding to the interference occupied period, preset empty symbol values ​​are filled in, ensuring that the subcarriers within the interference occupied period do not carry valid data. The signal generation module arranges the complex symbols of all subcarriers within the same symbol period according to the subcarrier index order to form an OFDM symbol vector. An inverse fast Fourier transform operation is performed on the OFDM symbol vector to generate a time-domain OFDM baseband signal. In practice, for symbol period 1, only subcarriers 1 and 3 are idle in the available subcarrier set, while subcarrier 4 is occupied. Therefore, the constructed orthogonal frequency division multiplexing symbol vector is filled with valid complex symbols at subcarrier indices 1 and 3, with an empty symbol value of 0 at subcarrier index 4, and 0 is also filled in other unused subcarrier index positions.

[0051] In some embodiments, the specific process of the inverse fast Fourier transform operation includes extracting the complex symbol sequence of all subcarriers in the current symbol period from the orthogonal frequency division multiplexing symbol vector, calculating the length of the complex symbol sequence, and comparing it with a preset number of fast Fourier transform points. If the length of the complex symbol sequence is equal to the preset number of fast Fourier transform points, the complex symbol sequence is directly used as the input sequence. If the length of the complex symbol sequence is less than the preset number of fast Fourier transform points, a corresponding number of zero-value symbols are added to the high-frequency positions of the complex symbol sequence to make the length of the input sequence reach the preset number of fast Fourier transform points. The input sequence is then input into the inverse fast Fourier transform calculation module in subcarrier index order. It can be understood that the preset number of fast Fourier transform points is 1024, while the actual number of subcarriers used is 600. Therefore, 424 zero-value symbols need to be added to the high-frequency part of the complex symbol sequence to form an input sequence of length 1024.

[0052] The inverse fast Fourier transform (IFFT) calculation module performs an inverse discrete Fourier transform (DFT) on the input sequence to generate an initial baseband signal sequence in the time domain. It then performs a cyclic prefix insertion operation on this initial baseband signal sequence, truncating a specified length of sample points from the end of the sequence and copying them to the beginning. After inserting the cyclic prefix, a windowing filter is applied to the resulting extended sequence to suppress out-of-band spectral leakage, generating the final time-domain orthogonal frequency division multiplexing (OFDM) baseband signal. The windowing filter operation is implemented by multiplying each sample point of the extended sequence by a window function coefficient. It can be calculated using the formula:

[0053] in: This represents the window function coefficient at sample point index n. This represents the index of the sample point in the extended sequence, with values ​​ranging from 0 to... , Indicates the length of the inserted cyclic prefix. This indicates the number of Fast Fourier Transform (FFT) points. Optionally, the length of the cyclic prefix can be set to one-quarter of the FFT points. It can be understood that windowing filtering effectively reduces out-of-band radiation by smoothly truncating the signal in the time domain. The signal generation module sequentially performs parallel-to-serial conversion, digital-to-analog conversion, and RF up-conversion on the time-domain orthogonal frequency division multiplexed baseband signal to generate a multi-carrier modulated transmission signal. Optionally, the RF up-conversion process shifts the baseband signal to a specified center frequency for transmission.

[0054] In one embodiment of the present invention, at the receiving end, the transmitted multi-carrier modulated signal is processed. The receiving end acquires a set of broadband spectrum data after down-conversion, analog-to-digital conversion, and serial-to-parallel conversion. A Fast Fourier Transform (FFT) operation is performed on this set, and the received complex symbols corresponding to each subcarrier in each symbol period are extracted from the transform result. Before further processing, the receiving end performs channel estimation and equalization. From the received complex symbols, the pilot symbol sequence at the pilot subcarrier positions preset by the transmitting end is extracted. This sequence is multiplied symbol-by-symbol by a known reference pilot symbol sequence stored locally at the receiving end to obtain the channel frequency response estimate at each pilot subcarrier position. Frequency domain interpolation is performed on these discrete channel estimates to estimate the channel frequency response of all subcarriers in all symbol periods. Using these channel frequency response estimates, a frequency domain equalization operation, such as zero-forcing equalization or minimum mean square error equalization, is performed on the original received complex symbols of each subcarrier in each symbol period to generate equalized received complex symbols after channel distortion compensation.

[0055] For each subcarrier, based on the modulation order configured by the transmitter for that subcarrier, soft demodulation is performed on the equalized received complex symbols to calculate the log-likelihood ratio of each bit, generating a soft bit metric sequence for that subcarrier. Then, based on the coding rate configured by the transmitter for that subcarrier, the corresponding channel decoding operation, such as Viterbi decoding or belief propagation decoding, is performed on the soft bit metric sequence to recover the decoded bit sequence carried by that subcarrier. Finally, the receiver concatenates and combines the decoded bit sequences of all subcarriers according to the data allocation order agreed upon by the transmitter during signal generation, thereby completely recovering the original data stream to be transmitted.

[0056] In practice, after generating the multi-carrier modulated transmission signal, the receiving end processes the multi-carrier modulated transmission signal to obtain the received wideband spectrum data set. Specifically, the receiving end performs serial-to-parallel conversion and Fast Fourier Transform (FFT) operations on the received wideband spectrum data set. The receiving end performs serial-to-parallel conversion on the wideband time-domain sampling sequence obtained through down-conversion and analog-to-digital conversion, dividing the long sequence into multiple equal-length data blocks. The length of each data block is equal to the sum of the number of FFT points used by the transmitting end and the length of the cyclic prefix. The receiving end removes the cyclic prefix portion at the beginning of each data block and performs FFT operations on the remaining data points to extract the received complex symbols of each subcarrier in each symbol period. In some embodiments, the transmitter uses 1024 Fast Fourier Transform points and a cyclic prefix length of 256. The receiver divides the received wideband spectrum data set into continuous data blocks of length 1280, removes the first 256 sample points of each data block, and performs a 1024-point Fast Fourier Transform on the remaining 1024 sample points. Each output point of the transform result corresponds to a received complex symbol of a subcarrier in one symbol period. Assuming that the system uses 512 subcarriers to transmit data, the receiver extracts the received complex symbols corresponding to these 512 subcarriers from the 1024 transform results.

[0057] After extracting the received complex symbols for each subcarrier in each symbol period, the receiver performs the following operation: The receiver extracts the pilot symbol sequence located at preset pilot subcarrier positions from the received complex symbols. The receiver performs a symbol-by-symbol conjugate multiplication operation with the locally stored reference pilot symbol sequence to obtain the estimated channel frequency response value at each pilot symbol position. In a specific implementation, the receiver knows that pilot symbols have been inserted on the subcarrier index set P = {10, 30, 50, ..., 990}, and extracts the symbols at these index positions from the received complex symbols to form the pilot symbol sequence. Locally stored reference pilot symbol sequence It is a known complex sequence pre-agreed upon by the transmitting end, usually using constant mode modulation, and the receiving end performs operations on each pilot position k. ,in This represents the estimated channel frequency response obtained at pilot subcarrier index k. This represents the complex symbol corresponding to index k in the received pilot symbol sequence. This represents the complex symbol corresponding to index k in the locally stored reference pilot symbol sequence. Indicates to The receiver performs frequency domain interpolation on the channel frequency response estimate to generate channel frequency response estimates for all subcarriers in each symbol period. Using the channel frequency response estimate, the receiver performs frequency domain equalization on the received complex symbols for each subcarrier in each symbol period, generating equalized received complex symbols. These equalized received complex symbols replace the original received complex symbols in the soft demodulation operation. It can be understood that the frequency domain equalization operation can employ a zero-forcing equalization algorithm. For a subcarrier index n and a symbol period t, the equalized received complex symbols... Calculated using the formula:

[0058] in: It is the original receiving complex number symbol. It is the channel frequency response estimate at this location obtained through frequency domain interpolation.

[0059] In some embodiments, the receiver presets 100 uniformly distributed pilot subcarriers, the reference pilot symbols are modulated using BPSK, and the pilot symbol sequence extracted by the receiver... Contains 100 complex numbers, and is related to the local reference sequence. After conjugate multiplication, 100 channel frequency response estimates are obtained. The receiving end uses a linear interpolation algorithm, utilizing these 100 discrete... The values ​​were used to estimate the channel frequency response for all 512 subcarrier locations used for data transmission. At the receiver, for each subcarrier, soft demodulation is performed on the equalized received complex symbols according to the modulation order corresponding to the subcarrier, generating a soft bit metric sequence for the subcarrier. The soft demodulation operation calculates the log-likelihood ratio of each bit carried by each complex symbol. For symbols using quadrature amplitude modulation, the soft bit metric can be obtained by calculating the Euclidean distance from the received symbol to each point in the constellation diagram. Channel decoding is then performed on the soft bit metric sequence according to the coding rate corresponding to the subcarrier, generating the decoded bit sequence of the subcarrier. The channel decoding operation, depending on the channel coding type used by the transmitter, such as Viterbi decoding or low-density parity-check decoding, uses the soft bit metric sequence to recover the uncoded information bit sequence.

[0060] The receiving end concatenates the decoded bit sequences of all subcarriers according to the data allocation order of the transmitting end to recover the original data stream to be transmitted. In specific implementations, the transmitting end allocates data according to the subcarrier index order and the distribution of interference idle periods during signal generation. The receiving end, based on the same subcarrier management information and interference status label sequence, knows which symbol periods each subcarrier carries valid data. The receiving end extracts the data blocks corresponding to the symbol periods from the decoded bit sequences of each subcarrier according to the subcarrier index order, and concatenates these data blocks in chronological order to finally recover the complete frame body data sequence. The frame body data sequence and the frame header synchronization sequence together constitute the recovered original data stream to be transmitted. Optionally, the data allocation order can be pre-agreed or informed to the receiving end through control channel signaling. It can be understood that the processing at the receiving end is the reverse process of the processing at the transmitting end, and its correct operation depends on the knowledge of the parameters and status of the transmitting end. In some embodiments, the available subcarrier set, the modulation order and coding rate of each subcarrier, and the interference status label sequence can be transmitted to the receiving end through an independent control link or preamble signal. The receiver performs the above-mentioned serial-to-parallel conversion, fast Fourier transform, channel estimation and equalization, soft demodulation, channel decoding and data splicing operations to reliably recover transmitted information in wireless channels with interference and fading.

[0061] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A multi-carrier broadband anti-interference communication system based on cognitive radio, characterized in that, The system includes: The spectrum sensing module acquires a set of broadband spectrum sampling data of the target communication frequency band within the current time window. The set of broadband spectrum sampling data includes the received signal strength indication value and noise floor estimate value corresponding to each of the multiple subcarriers. The interference identification module calls an improved hidden Markov model to perform interference state sequence decoding on the broadband spectrum sampling data set, generating an interference state label sequence for each subcarrier within the current time window. The improved hidden Markov model constructs a state transition constraint matrix based on the spectral correlation between adjacent subcarriers. The subcarrier management module selects subcarriers with idle interference status labels from the plurality of subcarriers based on the interference status label sequence, thereby forming a set of available subcarriers; The parameter configuration module performs adaptive modulation and coding parameter configuration processing on each subcarrier in the available subcarrier set to generate the modulation order and coding rate corresponding to each subcarrier. The signal generation module allocates the data stream to be transmitted to each subcarrier in the set of available subcarriers according to the modulation order and coding rate, thereby generating a multi-carrier modulated transmission signal. The improved Hidden Markov Model is invoked to perform interference state sequence decoding on the broadband spectrum sampling data set, generating an interference state label sequence for each subcarrier within the current time window, including: The received signal strength indication values ​​of each subcarrier at multiple sampling times within the current time window are extracted from the broadband spectrum sampling data set to form the observation time series of the subcarrier; The noise basis estimate of each subcarrier at multiple sampling times within the current time window is extracted from the broadband spectrum sampling data set to form the noise basis time series of the subcarrier; For each subcarrier, the ratio of the received signal strength indication value at each sampling time in the observation time series to the noise basis estimate value at the corresponding sampling time in the noise basis time series is calculated to generate the signal-to-noise ratio time series of the subcarrier; The signal-to-noise ratio time series of each subcarrier is input into the improved hidden Markov model as the observation sequence; The improved Hidden Markov Model performs Viterbi decoding on the observation sequence based on a preset state transition probability matrix, observation probability matrix, and initial state probability distribution. The off-diagonal elements in the state transition probability matrix are adaptively adjusted according to the correlation between the interference state labels of adjacent subcarriers. The Viterbi decoding operation outputs the interference state label that maximizes the joint probability at each sampling time. The interference state labels at all sampling times are arranged in chronological order to generate the interference state label sequence, wherein each interference state label is either occupied or idle.

2. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 1, characterized in that, The improved Hidden Markov Model constructs a state transition constraint matrix based on the spectral correlation between adjacent subcarriers, including: Obtain a subcarrier index sequence formed by arranging all subcarriers within the target communication frequency band in ascending order of frequency; For each subcarrier index in the subcarrier index sequence, determine the preceding and following subcarrier indices adjacent to the subcarrier index to form a set of adjacent subcarrier pairs; For each adjacent subcarrier pair, the co-occurrence frequency of two subcarriers in the adjacent subcarrier pair being occupied by the interference source at the same time, and the co-occurrence frequency of two subcarriers being in an idle state at the same time are counted from the historical spectrum monitoring data. The state consistency coefficient of adjacent subcarrier pairs is calculated based on the co-occurrence frequency. The state consistency coefficient is used to characterize the probability that the interference state labels of adjacent subcarrier pairs are the same. The state consistency coefficient is used as a weighting factor and multiplied by the corresponding element in the preset basic state transition probability matrix to generate the value of each element in the state transition constraint matrix. Normalize each row vector in the state transition constraint matrix so that the sum of all elements in each row vector equals the value one.

3. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 2, characterized in that, The adaptive adjustment process of the improved hidden Markov model includes: At the start of the current time window, the state transition probability matrix is ​​initialized to the state transition constraint matrix; After processing each sampling moment within the current time window, obtain the interference state labels obtained from decoding all subcarriers at the sampling moment to form a full-band state vector; The full-band state vector is scanned by a sliding window according to the subcarrier index order, and the label consistency ratio of adjacent subcarrier pairs is counted within the sliding window. The label consistency ratio is compared with a preset consistency threshold. When the label consistency ratio is lower than the consistency threshold, the values ​​of the diagonal elements in the state transition constraint matrix are reduced, while the values ​​of the off-diagonal elements in the state transition constraint matrix are increased. The adjusted state transition constraint matrix is ​​then normalized again to generate an updated state transition probability matrix, which is used for the Viterbi decoding operation at the next sampling time.

4. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 2, characterized in that, For each subcarrier in the available subcarrier set, an adaptive modulation and coding parameter configuration process is performed to generate the modulation order and coding rate corresponding to each subcarrier, including: Select a target subcarrier from the available subcarrier set, and extract the end value of the signal-to-noise ratio time series of the target subcarrier at the last sampling time in the current time window, as the current signal-to-noise ratio estimate of the target subcarrier; Obtain a mapping table between multiple preset signal-to-noise ratio ranges and modulation and coding schemes. Each modulation and coding scheme contains a modulation order value and a coding rate value. Traverse all signal-to-noise ratio (SNR) intervals in the mapping table, and determine the modulation and coding scheme corresponding to the SNR interval to which the current SNR estimate belongs as the candidate modulation and coding scheme for the target subcarrier; The interference status labels of the target subcarrier at all sampling times within the current time window are extracted from the interference status label sequence. The proportion of the sampling times with the interference status label being in the occupied state to the total number of sampling times in the current time window is counted as the interference duty cycle of the target subcarrier. The coding rate in the candidate modulation and coding scheme is adjusted downward based on the interference duty cycle to generate a corrected coding rate value, wherein the magnitude of the downward adjustment is positively correlated with the magnitude of the interference duty cycle. The modulation order of the target subcarrier is determined as the modulation order value in the candidate modulation and coding scheme, the coding rate of the target subcarrier is determined as the corrected coding rate value, and the operation is repeated for the next subcarrier in the set of available subcarriers until all subcarriers obtain the corresponding modulation order and coding rate.

5. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 1, characterized in that, The data stream to be transmitted is allocated to each subcarrier in the available subcarrier set according to the modulation order and coding rate, generating a multi-carrier modulated transmission signal, including: Obtain the original bit sequence of the data stream to be sent, and divide the original bit sequence into multiple data frames according to a preset frame structure. Each data frame contains a frame header synchronization sequence and a frame body data sequence. For each subcarrier in the set of available subcarriers, channel coding is performed on the frame data sequence to be carried by the subcarrier according to the coding code rate corresponding to the subcarrier to generate a coded bit sequence; A constellation mapping operation is performed on the encoded bit sequence according to the modulation order corresponding to the subcarrier to generate a complex symbol sequence corresponding to the subcarrier; Arrange the complex symbols of all subcarriers within the same symbol period in subcarrier index order to form an orthogonal frequency division multiplexing symbol vector; Perform an inverse fast Fourier transform operation on the orthogonal frequency division multiplexing symbol vector to generate a time-domain orthogonal frequency division multiplexing baseband signal; The time-domain orthogonal frequency division multiplexing baseband signal is sequentially subjected to parallel-to-serial conversion, digital-to-analog conversion, and radio frequency up-conversion to generate the multi-carrier modulated transmission signal.

6. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 5, characterized in that, Performing an inverse fast Fourier transform operation on the orthogonal frequency division multiplexing symbol vector to generate a time-domain orthogonal frequency division multiplexing baseband signal specifically includes: Extract the complex symbol sequence of all subcarriers in the current symbol period from the orthogonal frequency division multiplexing symbol vector; Calculate the length of the complex symbol sequence and compare it with the preset number of points in the Fast Fourier Transform; If the length of the complex symbol sequence is equal to the preset number of Fast Fourier Transform points, then the complex symbol sequence is directly used as the input sequence. If the length of the complex symbol sequence is less than the preset number of Fast Fourier Transform points, then a corresponding number of zero-value symbols are added to the high-frequency positions of the complex symbol sequence to make the length of the input sequence reach the preset number of Fast Fourier Transform points. The input sequence is input into the inverse fast Fourier transform calculation module in subcarrier index order; The inverse fast Fourier transform calculation module performs an inverse discrete Fourier transform operation on the input sequence to generate an initial baseband signal sequence in the time domain; A cyclic prefix insertion operation is performed on the initial baseband signal sequence in the time domain, truncating a specified length of sample points from the end of the sequence and copying them to the beginning of the sequence; After inserting the cyclic prefix, a windowing filter is applied to the resulting extended sequence to suppress out-of-band spectral leakage and generate the final time-domain orthogonal frequency division multiplexing baseband signal.

7. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 5, characterized in that, Before performing constellation mapping operation on the encoded bit sequence according to the modulation order corresponding to the subcarrier to generate the complex symbol sequence corresponding to the subcarrier, the method further includes performing: Extract the interference status label of each subcarrier at each sampling moment within the current time window from the interference status label sequence, and determine the interference idle time distribution of each subcarrier in the time dimension; Based on the distribution of the interference idle time periods, the encoded bit sequence corresponding to each subcarrier is divided into multiple sub-data blocks, and the length of each sub-data block is matched with the number of bits that the subcarrier can transmit in the corresponding idle time period; In the process of constructing the orthogonal frequency division multiplexing symbol vector, only the complex symbols in the complex symbol sequence that are located within the distribution of the interference idle time period are filled into the corresponding symbol period positions; For the symbol period position in the orthogonal frequency division multiplexing symbol vector corresponding to the interference-occupied period, a preset empty symbol value is filled in so that the subcarriers during the interference-occupied period do not carry valid data.

8. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 1, characterized in that, After generating the multi-carrier modulated transmission signal, the process further includes: The receiving end processes the multi-carrier modulated transmitted signal to obtain the received broadband spectrum data set; Perform serial-to-parallel conversion and fast Fourier transform operations on the received broadband spectrum data set to extract the received complex symbols of each subcarrier in each symbol period; For each subcarrier, a soft demodulation operation is performed on the received complex symbols according to the modulation order corresponding to the subcarrier to generate a soft bit metric sequence for the subcarrier; The soft bit metric sequence is subjected to channel decoding based on the coding rate corresponding to the subcarrier to generate the decoded bit sequence of the subcarrier; The decoded bit sequences of all subcarriers are concatenated according to the data allocation order at the transmitting end to recover the original data stream to be transmitted.

9. The multi-carrier broadband anti-interference communication system based on cognitive radio according to claim 8, characterized in that, After performing serial-to-parallel conversion and fast Fourier transform operations on the received broadband spectrum data set, and extracting the received complex symbols of each subcarrier in each symbol period, the process further includes: Extract the pilot symbol sequence located at the preset pilot subcarrier position from the received complex symbols; The pilot symbol sequence is multiplied symbol by symbol by symbol by the locally stored reference pilot symbol sequence to obtain the channel frequency response estimate at each pilot symbol position; Perform frequency domain interpolation on the channel frequency response estimate to generate channel frequency response estimates for all subcarriers in each symbol period; The channel frequency response estimate is used to perform frequency domain equalization on the received complex symbols of each subcarrier in each symbol period to generate equalized received complex symbols, which are used to replace the original received complex symbols in the soft demodulation operation.

Images