A time-frequency hybrid MMSE-NP-RISIC equalization method

Abstract

This invention belongs to the field of wireless communication and relates to a time-frequency hybrid MMSE-NP-RISIC equalization method, including: S1, acquiring the signal y at the receiving end. n Then, MMSE equalization is performed to obtain the equalized signal Z. k S2, and the decision signal, perform residual inter-symbol interference estimation and noise interference estimation on the decision signal to obtain the residual inter-symbol interference factor and noise interference factor S3, based on the factors and the signal Z k The process involves obtaining the decision signal S4 for the current iteration t, determining whether the iteration stopping condition has been met, and if so, obtaining the decision signal; otherwise, returning to step S2. This invention performs noise prediction during the feedback process of the MMSE-RISIC algorithm, improving equalization accuracy. The noise prediction and estimation in this invention are independent of MMSE equalization, simplifying the iterative calculation process and improving equalization efficiency. This invention suppresses and eliminates noise and residual inter-symbol interference in the time and frequency domains respectively, improving equalization accuracy.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication and relates to a time-frequency hybrid MMSE-NP-RISIC equalization method. Background Technology

[0002] Multipath fading is one of the main bottlenecks affecting the development of high-speed wireless communication. It manifests as inter-symbol interference (ISI) in the time domain, which leads to signal distortion and thus affects the reliability and stability of the system.

[0003] Currently, technologies for mitigating multipath effects mainly fall into two categories: single-carrier transmission technology and multi-carrier transmission technology. In single-carrier transmission technology, SC-FDE is a key technology, offering advantages such as good multipath fading resistance, improved signal quality and stability, and reduced implementation complexity and cost. In multi-carrier transmission technology, OFDM technology is widely used. Compared to other multi-carrier technologies, OFDM has lower implementation complexity and stronger anti-interference and anti-fading capabilities.

[0004] Single-carrier frequency division multiplexing (SC-FDM) technology combines the advantages of OFDM and SC-FDE technologies while avoiding their shortcomings. SC-FDM frequency domain equalization methods mainly include zero-forcing (ZF) equalization and minimum mean square error (MMSE) equalization; however, both of these methods are linear equalizations. To obtain better error performance, researchers often consider using nonlinear equalization methods, such as iterative equalization or equivalent decision feedback equalization (DFE), to eliminate ISI. In time-domain decision feedback equalization (TD-DFE), the feedforward filter is implemented in the time domain and feedback is performed on a symbol-by-symbol basis; the time-domain feedback filter is applied to previously detected symbols. Although TD-DFE offers better performance, its implementation complexity is high. To address this issue, the papers "On the comparison between OFDM and single carrier modulation with a DFE using a frequency-domain feedforward filter" and "Frequency domain equalization for single-carrier broadband wireless systems" proposed a hybrid decision feedback equalization (H-DFE) method, where the feedforward filter is implemented in the frequency domain while maintaining the time-domain feedback filter. Nevertheless, H-DFE still requires high complexity because the time-domain feedback filter must be optimized for the decision of each symbol. To further improve this, the paper "Single carrier frequency domain equalization with time domain noise prediction for wideband wireless communications" proposed a noise prediction decision feedback equalization (NP-DFE) structure, combining a linear frequency domain equalizer and a time-domain noise predictor, with a structure similar to H-DFE.The paper "A novel decision feedback equalizer for SC-FDE system" proposes an MMSE-RISIC algorithm, which uses the decision data after MMSE equalization to estimate and eliminate residual inter-symbol interference.

[0005] However, in the iterative steps of residual inter-symbol interference (ISI) elimination, the MMSE-RISIC algorithm suffers from the same amount of noise amplification as the residual ISI is reduced iteratively, resulting in poor equalization performance. Summary of the Invention

[0006] To address the aforementioned problems in the prior art, this invention employs a time-frequency hybrid MMSE-NP-RISIC equalization method, comprising:

[0007] S1. Obtain the signal y from the receiver of the SC-FDM system. n For signal y n Perform MMSE equalization to obtain the equalized signal Z. k and judgment signal

[0008] S2, the decision signal obtained from the previous iteration t-1. Residual inter-symbol interference (ISI) and noise interference (MI) estimations are performed to obtain the residual ISI factor. and noise interference factors

[0009] S3, Based on the residual inter-symbol interference factor and noise interference factors For the equalization signal Z k The process is performed to obtain the decision signal for the current iteration t.

[0010] S4. Determine whether the preset iteration stop condition has been met. If so, obtain the final decision signal. Otherwise, return to step S2;

[0011] Among them, SC-FDM is single-carrier frequency domain equalization, MMSE is minimum mean square error, and NP is noise prediction.

[0012] Beneficial effects:

[0013] 1. To address the shortcoming of traditional MMSE-RISIC algorithms that do not consider the impact of noise, the MMSE-NP-RISIC equalization method of this invention performs frequency domain noise prediction during the feedback process, thereby improving equalization accuracy. 2. In this invention, noise prediction, residual inter-symbol interference estimation, and MMSE equalization are all independent of each other. Therefore, it is not necessary to update the coefficients of the noise predictor and residual inter-symbol interference estimation during the iteration process, simplifying the iterative calculation process and improving equalization efficiency. 3. This invention is based on time-frequency hybrid hard-decision feedback equalization. Time-domain interference suppression can finely adjust the interference between symbols from a local perspective, while frequency-domain interference suppression can optimize the frequency components of the signal from a global perspective. The combination of the two enables the equalization to more accurately estimate and recover the original signal, thereby improving equalization accuracy. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of a single-carrier frequency division multiplexing system provided in an embodiment of the present invention;

[0015] Figure 2 This is a structural diagram of the MMSE-RISIC decision feedback equalizer provided in an embodiment of the present invention;

[0016] Figure 3 This is a diagram of the MMSE-NP-RISIC equilibrium structure provided in an embodiment of the present invention;

[0017] Figure 4 This is a performance comparison diagram of three equalization algorithms under Rayleigh channel provided in an embodiment of the present invention;

[0018] Figure 5 A schematic diagram illustrating the convergence of the MMSE-NP-RISIC method under a good shortwave channel provided in this embodiment of the invention;

[0019] Figure 6 A schematic diagram showing the performance comparison of the MMSE-NP-RISIC method provided in this embodiment of the invention under three shortwave environments;

[0020] Figure 7 This is a schematic diagram showing the performance comparison of three equalization algorithms under adverse channel conditions provided in an embodiment of the present invention.

[0021] Figure 8 This is a schematic diagram showing the performance comparison of three equalization algorithms under good channel conditions, provided in an embodiment of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] SC-FDM is a single-carrier modulation technique with frequency-domain equalization at the receiver and allows for the parallel transmission of data from multiple users. It is a variant of OFDM with additional FFT and IFFT processing blocks at the transmitter and receiver, respectively.

[0024] like Figure 1 As shown, at the transmitting end of the SC-FDM system, the input data stream is first modulated to convert it into a form suitable for transmission. Next, serial-to-parallel conversion transforms the data stream into multiple parallel data streams, which are then used... M The point-based Fast Fourier Transform (FFT) converts the signal from the time domain to the frequency domain and performs subcarrier mapping, mapping data symbols onto different subcarriers for transmission. Subsequently, an N-point Inverse Fast Fourier Transform (IFFT) is performed (where N > M), and a cyclic prefix (CP) is added. Finally, the signal is converted from a digital signal to an analog signal via a digital-to-analog converter (DAC) for transmission over a physical channel.

[0025] At the receiver, the analog signal is first converted back to a digital signal using an analog-to-digital converter (ADC). After removing the cyclic prefix (CP), an N-point FFT transform is performed to convert the SC-FDM signal to the frequency domain. In the frequency domain, subcarrier inverse mapping and frequency equalization are performed. Frequency equalization effectively compensates for multipath effects in the channel and mitigates the impact of frequency-selective fading caused by inter-symbol interference (ISI) on the system. Next, an M-point IFFT transforms the signal back to the time domain. Finally, parallel-to-serial conversion and decision processing are performed to recover the data transmitted by the transmitter.

[0026] In SC-FDM systems, there are two main subcarrier allocation methods. The first is centralized mapping, which maps all user data onto contiguous subcarriers, forming a single frequency domain block. The second is distributed mapping, which divides the bandwidth into a set of non-contiguous subcarrier blocks and maps each user's data onto one of these subcarrier sets, thus achieving diversity in the frequency domain. A special case of distributed mapping is interleaved mapping, which divides the bandwidth at the maximum interval and maps user data onto these divided subcarriers to achieve diversity in both the time and frequency domains. After subcarrier allocation, an N-point inverse FFT (IFFT) is used to transform the signal to the time domain.

[0027] In an SC-FDM system, the total number of users equals the bandwidth spreading factor Q = N / M, where N is the total number of subcarriers. This invention will transmit the user data block tail L before transmission. cp Each symbol is copied to the data block header to form a cyclic prefix (CP), which is typically removed before any major processing. If the CP length is the same as or longer than the multipath channel delay spread, it helps prevent inter-block interference (IBI) and also ensures orthogonality between different subcarriers, thus avoiding interference between subcarriers.

[0028] like Figure 2 As shown, the MMSE-RISIC algorithm uses the decision data after MMSE equalization to estimate and eliminate residual inter-symbol interference (ISI). However, in the iterative steps of residual ISI elimination, the MMSE-RISIC algorithm suffers the same amount of noise amplification as the residual ISI is reduced iteratively. Therefore, based on the above-mentioned SC-FDM system, this invention adopts a time-frequency hybrid MMSE-NP-RISIC equalization method, as follows: Figure 3 As shown, it includes:

[0029] S1. Obtain the signal y from the receiver of the SC-FDM system. n For signal y n Perform MMSE equalization to obtain the equalized signal Z. k and judgment signal

[0030] Assume the transmitted signal from the transmitter is x. n =[x0,x1,…,x (N-1) ] T After inserting the cyclic prefix CP, the signal is transmitted through a wireless channel. The receiver receives the transmitted signal and converts it back to a digital signal using a digital-to-analog converter. The cyclic prefix CP is then removed from the digital signal to obtain the signal y. n=[y0,y1,…,y (N-1) ] T Its time-domain expression can be written as:

[0031]

[0032] Among them, h n It is the channel impulse response matrix, v n It has a mean of 0 and a variance of σ. 2 Additive white Gaussian noise.

[0033] For signal y n Performing MMSE equalization includes:

[0034] S11, for signal y n Perform an FFT operation to obtain the frequency domain signal Y. k ;

[0035] Received signal y n After FFT transformation, it can be represented as:

[0036] Y k =H k X k +V k (2)

[0037] Among them, H k It is the channel impulse response matrix, X k For the transmit signal of the SC-FDM system's transmitter, V k It has a mean of 0 and a variance of σ. 2 Additive white Gaussian noise.

[0038] S12. Use the MMSE equalizer to adjust the frequency domain signal Y. k The process is performed to obtain the equalization signal Z. k ;

[0039] MMSE equalizer coefficients in, σ is the conjugate matrix of the channel impulse response matrix. 2 Let P be the noise power of the additive white Gaussian noise in the SC-FDM system, and let P be the power of the transmitted signal at the transmitter of the SC-FDM system.

[0040] From equation (2), we can see that the data signal after MMSE equalization is:

[0041] Z k =W k Y k (3)

[0042]

[0043] in

[0044]

[0045] Δ k =[Δ0,Δ1,…,Δ N-1 ] T The frequency domain form of the residual inter-symbol interference after equalization. This is frequency domain noise.

[0046] S13, Regarding the equalization signal Z k Perform IFFT operation to obtain the time-domain signal z. n ;

[0047] Applying an IFFT to equation (4) yields:

[0048]

[0049] In the formula, δ n and These are the residual inter-symbol interference and noise interference after MMSE equalization.

[0050] S14. For the time-domain signal z n Make a judgment and receive a judgment signal.

[0051] MMSE equilibrium directly for z n Making a decision can lead to inaccurate equalization data. Therefore, the MMSE-NP-RISIC equalization method of this invention obtains a more accurate decision by estimating and eliminating residual inter-symbol interference and noise interference.

[0052] S2, the decision signal obtained from the previous iteration t-1. Residual inter-symbol interference (ISI) and noise interference (MI) estimations are performed to obtain the residual ISI factor. and noise interference factors

[0053] For signal Residual inter-symbol interference estimation and noise interference estimation include:

[0054] S21, Regarding the signal Perform FFT operation to obtain the signal

[0055] S22, according to the signal Residual inter-symbol interference (ISI) is estimated to obtain the residual ISI factor.

[0056] Signal Substituting the frequency domain form of residual inter-symbol interference The estimated residual inter-symbol interference is obtained. right Perform IFFT operations to obtain the estimated residual inter-symbol interference factor.

[0057] As can be seen from the above process, the residual inter-symbol interference estimation and MMSE equalization are independent of each other. Therefore, it is not necessary to update the coefficients of the residual inter-symbol interference estimation during the iteration process, which simplifies the iteration calculation process and improves the equalization efficiency.

[0058] S23, according to the signal and signal Z k Noise interference is estimated to obtain the noise interference factor.

[0059] Signal and signal Z k Superposition yields the signal Calculate the coefficients of the noise predictor Signal With coefficient B k Multiply to obtain the noise interference factor. Among them, H m Let N be the channel impulse response matrix on the m-th subcarrier, where m is the index of the subcarrier in the SC-FDM system and N is the total number of subcarriers in the SC-FDM system.

[0060] The calculation of the coefficients for the noise predictor includes:

[0061] Assume R k =X k Then the error E k for:

[0062] E k =U k -X k =(1-B k (W) k H k -1)X k +(1-B k W k V k (8)

[0063] Construct the mean square error E[ε]:

[0064]

[0065] Introduce constraints:

[0066]

[0067] Among them, B mTo predict the noise coefficients on the m-th subcarrier, a cost function J is constructed using the Lagrange multiplier method:

[0068]

[0069] Where λ is the Lagrange multiplier, derived from...

[0070]

[0071] have to

[0072]

[0073] in, W is the coefficient of the MMSE equalizer. k conjugate, B is the coefficient of the noise predictor. k The conjugate of W can be seen. k With B k Irrelevant, substitute J, then by

[0074]

[0075] have to

[0076]

[0077] Then through

[0078]

[0079] have to

[0080]

[0081] but

[0082]

[0083] As can be seen from the formula, noise prediction and MMSE equalization are independent of each other, and only require the channel information H to be obtained. k Then the equalizer coefficient W k With noise predictor coefficient B k Therefore, the coefficients of the noise predictor do not need to be updated during the iteration process, which simplifies the iteration calculation process and improves the equalization efficiency.

[0084] S3, Based on the residual inter-symbol interference factor and noise interference factors For the equalization signal Z k The process is performed to obtain the decision signal for the current iteration t.

[0085] For the equalization signal Zk The processing includes:

[0086] S31, Equalize signal Z k and noise interference factors By superimposing the signals, a signal with noise interference eliminated is obtained. For signal Perform IFFT operation to obtain the time-domain signal.

[0087] S32, convert the time-domain signal and residual inter-symbol interference factor Superposition yields a signal with residual inter-symbol interference eliminated. For signal Make a judgment and receive a judgment signal.

[0088] As can be seen from the above process, the present invention is based on hard decision feedback equalization with time-frequency hybridization, which suppresses and eliminates noise and residual inter-symbol interference in the time domain and frequency domain respectively, thereby improving the equalization accuracy.

[0089] S4. Determine whether the preset iteration stop condition has been met. If so, obtain the final decision signal. Otherwise, return to step S2.

[0090] The maximum number of iterations is D. When the number of iterations reaches the maximum number of iterations, the stopping condition is met. Preferably, D is 2.

[0091] In the MMSE-NP-RISIC equalization method of this invention, noise prediction, δ estimation, and MMSE equalization are independent of each other, and the method supports an iterative process. Therefore, it is not necessary to update the coefficients of the noise predictor and residual inter-symbol interference during the iteration process, which simplifies the iterative calculation process to some extent. However, the introduction of iteration also increases the computational load, so whether to adopt an iterative process should be determined based on the accuracy requirements.

[0092] Table 1 presents a comparative analysis of the complexity of several different equalization algorithms, where the number of FFT transformation points N represents the algorithm's complexity. The MMSE-NP-RISIC algorithm adds noise suppression processing, resulting in a slightly higher complexity than the MMSE-RISIC algorithm.

[0093] Table 1 Comparison of the complexity of various load balancing algorithms

[0094]

[0095] To verify the effectiveness of the proposed MMSE-NP-RISIC equalization method in shortwave communication, this invention simulates and compares several equalization algorithms. In the simulation, a shortwave channel under mid-latitude quiet conditions is used as the channel model, and an SC-FDM system is used as the communication model. Ideal channel synchronization is assumed, and compressed sensing (CS) channel estimation is employed. The relevant system parameters are shown in Table 2.

[0096] Table 2 System Simulation Parameters

[0097]

[0098]

[0099] Figure 4 Performance comparisons of three equalization algorithms were performed under Rayleigh channels (iturHFMQ / rayleigh). The MMSE-RISIC algorithm and the MMSE-NP-RISIC algorithm proposed in this invention both used a set number of iterations of 2. Figure 4 It can be clearly observed that the improved algorithm shows a significant performance improvement over the MMSE-RISIC equalization algorithm under Rayleigh channel conditions, with a bit error rate of 10%. -4 At this point, it achieves a performance gain of 0.9 dB compared to the MMSE-RISIC algorithm. Although the complexity of the MMSE-NP-RISIC algorithm proposed in this invention is slightly higher than that of the MMSE-RISIC algorithm, it achieves good performance by sacrificing a small amount of complexity.

[0100] Table 3 provides a brief introduction and analysis of shortwave channels in three different environments from the perspectives of multipath delay spread and Doppler spread. iturHFMQ, iturHFMM, and iturHFMD are a series of standard channel models defined by the International Telecommunication Union Radiocommunication Sector (ITU-R).

[0101] Table 3 Shortwave Channel Parameters

[0102]

[0103] Figure 5 This paper examines the convergence of the proposed MMSE-NP-RISIC algorithm during the iterative process under favorable shortwave conditions. Figure 5 As shown, the algorithm has essentially converged after two iterations. The theoretical part shows that the method proposed in this invention is equivalent to the MMSE equalization algorithm with zero iterations, and achieves good results with only one iteration, at a bit error rate of 10%. -4 A gain of 2.1dB can be obtained at that time.

[0104] Figure 6 This paper compares the performance of the MMSE-NP-RISIC algorithm proposed in this invention under three different shortwave channel environments. The MMSE-NP-RISIC algorithm uses a set number of iterations of 2. From... Figure 6 The comprehensive comparison shows that the equalization algorithm can fully leverage its advantages and effectively reduce the bit error rate under good channel conditions. However, in moderate or poor channel environments, the equalization effect weakens due to increased channel fading and interference, resulting in a higher bit error rate. Therefore, the better the channel conditions, the better the performance of the equalization algorithm.

[0105] Figure 7 This section compares the performance of the various equalization algorithms mentioned above under adverse channel conditions. The MMSE-RISIC algorithm and the MMSE-NP-RISIC algorithm proposed in this paper both have an iteration count of 2. From... Figure 7 As can be seen, under adverse channel conditions, the algorithm proposed in this invention exhibits a significant performance improvement compared to other algorithms. In the low signal-to-noise ratio (SNR) region, the improvement is not significant, but as the SNR increases, the advantages of the improved algorithm gradually become apparent, with a significant reduction in the bit error rate, especially under high SNR conditions, where its advantages are even more pronounced.

[0106] Figure 8 This section compares the performance of the various equalization algorithms described above under favorable channel conditions. The MMSE-RISIC algorithm and the MMSE-NP-RISIC algorithm proposed in this paper both use a set number of iterations of 2. From... Figure 8 It can be seen that when the bit error rate is 10 -4 The proposed MMSE-NP-RISIC algorithm improves performance by approximately 0.8 dB compared to the standard MMSE-RISIC algorithm and by 1.6 dB compared to the traditional MMSE algorithm. This performance improvement stems from the enhanced noise handling capabilities of the MMSE-NP-RISIC algorithm, effectively suppressing residual inter-symbol interference (ISI). This algorithm not only effectively reduces the impact of noise on signal quality but also minimizes residual ISI by precisely adjusting and optimizing algorithm parameters, thereby improving the overall system performance.

[0107] In summary, this invention first analyzes the shortcomings of the MMSE-RISIC equalization algorithm in shortwave communication within the SC-FDM system, particularly its failure to adequately consider the impact of noise on the signal during MMSE equalization feedback, resulting in unsatisfactory equalization performance. To address this issue, a time-frequency hybrid MMSE-NP-RISIC equalization method is employed. To overcome the limitation of the traditional MMSE-RISIC algorithm in neglecting noise, the improved structure introduces a frequency-domain noise predictor in the feedback branch and incorporates time-domain residual inter-symbol interference cancellation technology, thereby enhancing equalization accuracy. This improvement achieves a significant performance boost while maintaining low computational complexity. Simulation results demonstrate that the MMSE-NP-RISIC algorithm outperforms the comparable MMSE-RISIC algorithm in equalization performance and exhibits superior performance compared to the traditional MMSE algorithm.

[0108] The above-described embodiments further illustrate the purpose, technical solution, and advantages of the present invention. It should be understood that the above-described embodiments are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made to the present invention within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A time-frequency hybrid MMSE-NP-RISIC equalization method, characterized in that, include: S1. Obtain the signal y from the receiver of the SC-FDM system. n For signal y n Perform MMSE equalization to obtain the equalized signal Z. k and judgment signal S2, the decision signal obtained from the previous iteration t-1. Residual inter-symbol interference (ISI) and noise interference (MI) estimations are performed to obtain the residual ISI factor. and noise interference factors On the judgment signal Residual inter-symbol interference estimation and noise interference estimation include: For signal Perform FFT operation to obtain the signal According to the signal Residual inter-symbol interference (ISI) is estimated to obtain the residual ISI factor. Using a noise predictor based on the signal and signal Z k Noise interference is estimated to obtain the noise interference factor. According to the signal and signal Z k Noise interference estimation includes: [The sentence is incomplete and requires more context to translate accurately.] and signal Z k Superposition yields the signal Calculate the coefficients of the noise predictor According to the signal With coefficient B k Obtain noise interference factor in, σ is the conjugate matrix of the channel impulse response matrix. 2 H represents the noise power of the SC-FDM system, P represents the power of the transmitted signal at the transmitter of the SC-FDM system, and H represents the noise power. m Let N be the channel impulse response matrix on the m-th subcarrier, where m is the index of the subcarrier in the SC-FDM system and N is the total number of subcarriers in the SC-FDM system. Calculating the coefficients of the noise predictor includes: constructing the mean square error E[ε] and introducing constraints. The cost function is constructed using the Lagrange multiplier method. Estimate the coefficients of the MMSE equalizer based on the cost function J, and estimate the coefficients B of the noise predictor based on the estimated coefficients of the MMSE equalizer and the cost function J. k Where λ is the Lagrange multiplier, and B m Let be the noise prediction coefficient on the m-th subcarrier; S3, Based on the residual inter-symbol interference factor and noise interference factors For the equalization signal Z k The process is performed to obtain the decision signal for the current iteration t. S4. Determine whether the preset iteration stop condition has been met. If so, obtain the final decision signal. Otherwise, return to step S2; Among them, SC-FDM is single-carrier frequency domain equalization, MMSE is minimum mean square error, and NP is noise prediction.

2. The time-frequency hybrid MMSE-NP-RISIC equalization method according to claim 1, characterized in that, For signal y n Performing MMSE equalization includes: S11, for signal y n Perform an FFT operation to obtain the frequency domain signal Y. k ; S12. Use the MMSE equalizer to adjust the frequency domain signal Y. k The process is performed to obtain the equalization signal Z. k ; S13, Regarding the equalization signal Z k Perform IFFT operation to obtain the time-domain signal z. n ; S14. For the time-domain signal z n Make a judgment and receive a judgment signal.

3. The time-frequency hybrid MMSE-NP-RISIC equalization method according to claim 1, characterized in that, Acquire the signal y at the receiver of the SC-FDM system n This includes: the receiver of an SC-FDM system receives the signal transmitted through the channel, converts the transmitted signal back to a digital signal through digital-to-analog conversion, removes the cyclic prefix (CP) from the digital signal, and obtains the signal y. n .

4. The time-frequency hybrid MMSE-NP-RISIC equalization method according to claim 1, characterized in that, According to the signal Estimating residual inter-symbol interference includes: Calculate signal Z k The frequency domain form of residual inter-symbol interference Δ k , will signal Substitute Δ k The estimated residual inter-symbol interference is obtained. right Perform IFFT operations to obtain the estimated residual inter-symbol interference factor.

5. The time-frequency hybrid MMSE-NP-RISIC equalization method according to claim 4, characterized in that, Calculate signal Z k The frequency domain form of residual intersymbol interference includes: coefficients of the MMSE equalizer. Frequency domain signal Y k =H k X k +V k Then the equalization signal Among them, H k The channel impulse response matrix, σ is the conjugate matrix of the channel impulse response matrix. 2 X is the noise power of the SC-FDM system, P is the power of the transmitted signal at the transmitter of the SC-FDM system, and X is the noise power. k For the transmit signal of the SC-FDM system's transmitter, V k It is additive white Gaussian noise. For frequency domain noise, For signal Z k The frequency domain form of residual inter-symbol interference.

6. The time-frequency hybrid MMSE-NP-RISIC equalization method according to claim 1, characterized in that, Based on residual inter-symbol interference factor and noise interference factors For the equalization signal Z k The processing includes: S31, Equalize signal Z k and noise interference factors By superimposing the signals, a signal with noise interference eliminated is obtained. For signal Perform IFFT operation to obtain the time-domain signal. S32, convert the time-domain signal and residual inter-symbol interference factor Superposition yields a signal with residual inter-symbol interference eliminated. For signal Make a judgment and receive a judgment signal.

7. The time-frequency hybrid MMSE-NP-RISIC equalization method according to claim 1, characterized in that, The determination of whether the preset iteration stopping condition has been met includes: when the number of iterations t reaches the preset maximum number of iterations, the iteration stopping condition is met.

Images