Random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method and system

Abstract

The invention discloses a random pre-coding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method and system, a transmitting end enables each symbol to fully utilize all sub-channels through random unitary transformation and random compression transformation, and at a receiving end, by means of a block diagonal channel matrix of MIMO-OFDM on a frequency domain, all sub-channels of the MIMO-OFDM are fully utilized, so that the random pre-coding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication is realized. A cross-domain orthogonal approximate information transfer (CD-OAMP) detector is designed, which can perform signal detection with extremely low complexity, and can be optimal for replica maximum posteriori (MAP). The RP-MIMO-OFDM performance of the CD-OAMP detector exceeds the performance of MIMO-OFDM, the performance of MIMO-OTFS and the performance of MIMO-AFDM, the performance of RP-MIMO-OFDM is close to the performance of MIMO-IFDM, and the complexity is lower.

Description

Technical Field

[0001] The present invention relates to the field of wireless communications, and in particular to a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication system. Background Art

[0002] Selective fading of channels is an important problem in wireless communications. For example, multipath fading is a typical channel fading. It refers to a channel in which the signal reaches the receiving end along multiple paths due to reflection, refraction, scattering and other effects of obstacles such as terrain, buildings and vegetation during transmission. The signals of these different paths are superimposed on each other at the receiving end, causing changes in the signal amplitude and phase, thereby producing fading phenomena. This channel has the characteristics of delay spread, frequency selective fading, Rayleigh fading and frequency flat fading. Delay spread causes the signal to expand in time, resulting in inter-symbol interference; frequency selective fading refers to the signal components of different frequencies experiencing different degrees of fading in the channel; in the Rayleigh fading channel, the amplitude and phase of the signal will change randomly, and the change in amplitude follows the Rayleigh distribution; and frequency flat fading means that the channel has the same fading characteristics for signal components of all frequencies.

[0003] Multipath fading channels have a significant impact on communication systems. They reduce signal quality, cause random changes in signal amplitude and phase, lead to signal distortion, reduce the signal-to-noise ratio, and affect the performance of communication systems. Inter-symbol interference caused by delay spread makes it difficult for the receiver to accurately demodulate the signal, increasing the bit error rate. In addition, the frequency selective fading characteristics of multipath fading channels will also reduce the capacity of the channel, limiting the data transmission rate of the communication system.

[0004] To address the challenges posed by multipath fading channels, communication systems have adopted various measures. Channel coding adds redundant information by encoding signals at the transmitter, enabling the receiver to correct errors caused by multipath fading to a certain extent and improving signal reliability. Diversity techniques utilize the independent fading characteristics of signals in different spatial, temporal, frequency, or polarization states. By receiving and combining multiple independently faded signal components, the impact of signal fading is reduced, and signal quality is improved. Adaptive equalization technology uses an adaptive equalizer at the receiver to adjust and compensate for signals in real-time according to channel characteristics, eliminating inter-symbol interference and restoring the original waveform of the signal. OFDM technology divides the signal into multiple subcarriers for transmission. Each subcarrier has a narrow bandwidth, which can convert a frequency-selective fading channel into multiple frequency-flat fading channels. By using appropriate modulation and demodulation methods on each subcarrier, the impact of multipath fading can be effectively resisted. MIMO technology utilizes the spatial diversity characteristics of multiple antennas to transmit and receive multiple independent data streams, improving signal transmission quality and channel capacity. In a multipath fading channel, it can make full use of multipath components to reduce the impact of signal fading. MIMO-OFDM is a wireless communication technology that combines multiple-input multiple-output (MIMO) technology and orthogonal frequency-division multiplexing (OFDM) technology. It uses multiple antennas at the transmitter for data transmission and divides the data into multiple sub-streams for transmission using OFDM technology, which can overcome the problems caused by multipath fading channels to a certain extent. However, it still faces sparse channels and fails to completely overcome channel selective fading.

[0005] Consider a single-antenna wideband OFDM symbol passing through a quasi-static multipath channel with a channel impulse response of h[n] and additive white Gaussian noise of w[n]. Assume the number of subcarriers of OFDM is K, and the time-domain OFDM symbol of length K is x[n]. Then the received signal r[n] is

[0006]

[0007] where P is the number of channel taps. Without loss of generality, it is generally assumed that P < K. A cyclic prefix (CP) is added before transmission to avoid inter-symbol interference. After the receiver removes the CP, the received signal model can be written as

[0008] y = Hx + w,

[0009] where w is white Gaussian noise with variance σ 2 , and the channel matrix H satisfies a cyclic shift structure and can thus be decomposed as where Λ = diag(λ 1 , …, λ K ) is the frequency-domain response of the channel, and F K is the normalized K-point fast Fourier transform (FFT) matrix.

[0010] For a tx transmit antennas and N rx A MIMO-OFDM system with N receiving antennas and a length of N T =KN tx The emission symbol vector Modulated to

[0011]

[0012] in represents the Kronecker product. Accepting signal Where N R =KN rx ,for

[0013]

[0014] in For nth r The nth transmitting antenna and t The multipath channel between the receiving antennas can be decomposed into in Then the signal received by each antenna is demodulated, that is,

[0015]

[0016] Therefore, the received symbol in the frequency domain It can be expressed as

[0017]

[0018] Frequency domain equivalent channel matrix It is still a sparse fading channel matrix. Although MIMO multi-antenna technology is used to resist certain frequency selective fading, it is still impossible to completely overcome channel fading, especially when the number of antennas is small. In summary, the MIMO-OFDM communication system still faces the following channel fading conditions: First, frequency selective fading. OFDM fights against frequency selective fading by dividing the broadband signal into multiple narrowband subcarriers, but on deep fading subcarriers, the signal may still be severely attenuated, resulting in an increase in bit error rate. Second, time selective fading. In high-speed mobile scenarios, the channel changes rapidly, and OFDM symbols may be affected by time selective fading, resulting in inter-subcarrier interference (ICI), affecting system performance. Third, spatial correlation. MIMO relies on multipath propagation to achieve spatial diversity, but when the antenna spacing is insufficient or the scattering environment is not rich, the spatial correlation is enhanced, the diversity effect is reduced, and it is difficult to completely eliminate channel fading. Summary of the invention

[0019] The present invention aims to address the deficiencies of the prior art and to propose a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method and system.

[0020] The object of the present invention is to achieve the following technical solution: a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method, the method comprising:

[0021] The transmitter performs random precoding on the symbol vector to be transmitted, and transmits the randomly coded signal to the receiver using the antenna after OFDM modulation;

[0022] The receiving antenna performs OFDM demodulation on the received signal and inputs it into a low-complexity cross-domain orthogonal approximate message passing receiver to construct a signal estimation problem containing random transformation prior information;

[0023] The low-complexity cross-domain orthogonal approximate message passing receiver includes a linear estimator and a nonlinear estimator;

[0024] The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, inputs the output of the linear estimator into the nonlinear estimator for symbol-by-symbol MMSE demodulation after inverse random transformation, and outputs the output of the nonlinear estimator as a signal detection result after the iteration is completed.

[0025] Furthermore, the random precoding matrix construction method used in the random precoding includes random unitary transformation or random compression transformation.

[0026] Furthermore, the random unitary transformation or random compression transformation specifically includes:

[0027] Random precoding based on Haar distribution unitary matrix: Perform singular value decomposition on an independent and identically distributed Gaussian matrix G, that is, G = U∑V H , where U and V are the unitary matrices of the Haar distribution;

[0028] Random precoding based on fast transformation: random precoding matrix Ξ=ΠT N , where Π is a random permutation matrix, T N is the FT matrix, including the Fast Fourier Transform or Walsh-Hadamard Transform;

[0029] Random precoding based on large-scale Gaussian matrix: For a matrix composed of independent and identically distributed Gaussian elements The N×N square matrix is ​​composed of. When N→∞, the square matrix is ​​regarded as a unitary matrix;

[0030] Generalized Compressed Random Precoding: Using an incomplete shuffle matrix Π of size N×M (N<M) in Construct a generalized contraction random matrix, that is, Ξ=Πin U, where U is a random unitary matrix;

[0031] Random precoding based on interleaved block sparse fast transform: The random precoding based on interleaved block sparse fast transform of size N×M is a random unitary transform when N = M, and a random compression transform when N < M. The vector is decomposed into L blocks, and multiple low-dimensional fast transform matrices are interleaved, where N = LN s , M = LM s , that is, a sparse matrix with random characteristics is constructed, that is

[0032]

[0033] where is the M s point fast transform, Π l (l = 1,..., L) is the N s ×M s built-in permutation matrix, and is the outer permutation matrix of N×N; when Π l (l = 1,..., L) is a positive matrix, that is, N s = M s at this time, IBSFT is a random unitary transform, and when L = 1, it degenerates into random precoding based on fast transform.

[0034] Furthermore, the construction of the signal estimation problem containing random transform prior information specifically includes:

[0035] The received signal is

[0036]

[0037] where is the frequency-domain channel between the transmitting antenna n t and the receiving antenna n r ; and are row-column interleaved matrices, and the above formula is rewritten as

[0038]

[0039] where where is AWGN, and the distribution remains unchanged;

[0040] The MIMO multipath channel is independently decomposed by subcarriers, that is

[0041]

[0042] where represents represents Similarly;

[0043] Signal detection is optimized as an estimation problem with the following two constraints:

[0044]

[0045] Φ:s i ~P s ,

[0046] where k = 1, ..., K, the element s in s i Press P s Independent and identically distributed. When N rx =N tx When =1, A degenerates into a diagonal matrix.

[0047] Furthermore, the linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, specifically comprising:

[0048] The prior information from NLE and random transformation Considered as independent and identically distributed Gaussian distribution, linear minimum mean square error estimation is performed under constraint Γ. For all subcarrier indices k, k = 1, ..., K, from start,

[0049]

[0050] in

[0051]

[0052] Furthermore, the inverse random transformation specifically includes:

[0053] In the frequency domain Converted to random transform domain After inverse random unitary transformation and orthogonalization, the input of the nonlinear estimator is regarded as Gaussian observations of s, that is,

[0054] Furthermore, the nonlinear estimator performs symbol-by-symbol MMSE demodulation specifically as follows:

[0055] Nonlinear estimator φ t (·) Includes symbol-by-symbol MMSE demodulation and orthogonalization, random transformation domain The output of s is estimated to be

[0056]

[0057] Orthogonalization:

[0058]

[0059] Furthermore, when the precoding uses random compression transformation, the inverse random transformation should The elements dropped from the LE input Make up for it in

[0060]

[0061] in For supplementary items, only The items in are discarded, and the other items are zero.

[0062] On the other hand, the present invention also provides a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication system for implementing the method, the system comprising:

[0063] A transmitter module and a receiver module;

[0064] The transmitter module includes a transmitting antenna and a precoding submodule; the symbol vector to be transmitted is randomly encoded by the precoding submodule, and the randomly encoded signal is transmitted to the receiver module using the antenna after OFDM modulation;

[0065] The receiver module includes a receiving end antenna and a low-complexity cross-domain orthogonal approximate message passing receiver; the receiving end antenna performs OFDM demodulation on the received signal and inputs it into the low-complexity cross-domain orthogonal approximate message passing receiver to construct a signal estimation problem containing random transformation prior information;

[0066] The low-complexity cross-domain orthogonal approximate message passing receiver includes a linear estimator, a nonlinear estimator and a random transformation module;

[0067] The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, inputs the output of the linear estimator into the nonlinear estimator for symbol-by-symbol MMSE demodulation after inverse random transformation, and outputs the output of the nonlinear estimator as a signal detection result after the iteration is completed.

[0068] On the other hand, the specification of the present invention also provides a randomly precoded multiple-input multiple-output orthogonal frequency division multiplexing wireless communication device, including a memory and one or more processors, wherein the memory stores executable code, and when the processor executes the executable code, it implements the random precoded multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method.

[0069] Beneficial effects of the present invention:

[0070] The present invention exhibits strong anti-fading properties in quasi-static multipath fading channels. Compared with existing MIMO-OFDM, MIMO-OTFS, MIMO-AFDM and other technologies, the bit error rate of the present invention is lower under the same signal-to-noise ratio. Figure 5 At the same time, the present invention can adjust the size of the random precoding matrix for compression coding, thereby obtaining a higher code rate, such as Figure 6 shown. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 This is the block diagram of the RP-MIMO-OFDM system;

[0072] Figure 2 Schematic diagram of block diagonalization of MIMO multipath fading channel matrix;

[0073] Figure 3 is the block diagram of the CD-OAMP receiver algorithm for random unitary transformation RP;

[0074] Figure 4 The block diagram of the CD-OAMP receiver algorithm for the compressed IBSFT transform RP;

[0075] Figure 5 Comparison of bit error rate performance of different modulation precoding technologies for 2×2 MIMO system;

[0076] Figure 6 Comparison of bit error rate performance of different modulation symbol sets and compression rates for 2×2 RP-MIMO-OFDM system;

[0077] Figure 7 A schematic diagram of a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0078] The specific implementation modes of the present invention are further described in detail below with reference to the accompanying drawings.

[0079] like Figure 1 As shown, the present invention provides a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method, characterized in that the method includes:

[0080] S1. The transmitter performs random precoding on the symbol vector to be transmitted, and transmits the randomly encoded signal to the receiver using an OFDM modulated antenna;

[0081] In a communication transmitter, random precoding (RP) is a precoding defined by the following random transformation:

[0082]

[0083] wherein is a signal vector, is a random precoding matrix independent of the signal and the channel, is the precoded signal, which is a frequency-domain symbol in the RP-MIMO-OFDM system. When N = M, RP is a random unitary transform; when N < M, RP is a random compression transform. Generally speaking, RP has good Gaussian characteristics, that is, it can transform a certain vector into an asymptotically Gaussian vector.

[0084] RP-MIMO-OFDM is a new architecture that adds an RP operation before modulation in the original MIMO-OFDM system, as Figure 1 shown. Assume that the number of subcarriers is K, there are N tx antennas at the transmitter, and N rx antennas at the receiver. For the transmitted symbol vector perform an RP of size N T ×M T , where N T = KN tx is the total number of subcarriers of all transmitter antennas, that is wherein is the signal vector after RP. The precoded signal to be transmitted is divided into N tx segments. OFDM modulation and demodulation are performed on each antenna at the transmitter and the receiver respectively, that is:

[0085]

[0086] wherein, F K is a normalized K-point fast Fourier transform (FFT) matrix, respectively represent identity matrices of size N tx and N rx , represents the Kronecker product, x and y are the transmitted symbol sequence and the received symbol sequence respectively, is the frequency-domain signal at the receiver.

[0087] The present invention provides some construction methods for the random unitary precoding matrix Ξ. The random unitary precoding can be a random unitary transform or a random compression transform.

[0088] (1) Haar-Randomized Precoding (Haar-RP): The Haar-distributed unitary matrix has complete randomness and can thus be used for randomized precoding. However, the Haar-distributed unitary matrix is usually not sparse, which increases the complexity of both the randomized precoding and the receiver. A commonly used method to obtain the Haar-distributed unitary matrix is to perform singular value decomposition (SVD) on an independent and identically distributed Gaussian matrix G, i.e., G = UΣV H , where U and V are Haar-distributed unitary matrices.

[0089] (2) Fast-Transform-Based Randomized Precoding (FT-RP): The fast transform (FT) with random permutation exhibits great randomness at a relatively low complexity , i.e., Ξ = ΠT N , where Π is a random permutation matrix and T N is an FT matrix, such as the fast Fourier transform (FFT) or Walsh-Hadamard transform (WHT).

[0090] (3) Large-Scale Gaussian Matrix-Based Randomized Precoding (Gauss-RP): For an N×N square matrix composed of independent and identically distributed Gaussian elements , when N → ∞, according to the law of large numbers, this matrix can be regarded as a unitary matrix.

[0091] (4) Generalized Compressed Randomized Precoding (General Compressed RP): An incomplete scrambling matrix Π of size N×M (N < M) can be used in to construct a generalized compressed random matrix, i.e., Ξ = Π in U, where U is a random unitary matrix.

[0092] (5) Interleaved Block Sparse Fast-Transform-Based Randomized Precoding (IBSFT-RP): The recently proposed interleaved block sparse fast transform (IBSFT) provides a new RP scheme with lower complexity. The IBSFT-RP of size N×M can be either a random unitary transform (when N = M) or a random compressed transform (when N < M). It decomposes the vector into L blocks and interleaves multiple low-dimensional fast transform matrices (where N = LN s , M = LM s ), i.e., constructs a sparse matrix with random characteristics, i.e.,

[0093]

[0094] where is an M s -point fast transform (FT, such as FFT or WHT), and Π l (l = 1,..., L) is an N s × M sThe internal permutation matrix of is the N×N external permutation matrix. l (l=1,…,L) is a positive matrix, that is, N s =M s When L = 1, IBSFT is a random unitary transformation. In this case, the internal permutation matrix is ​​unnecessary. When L = 1, IBSFT-RP degenerates to FT-RP.

[0095] In summary, Haar-RP, FT-RP, and Gauss-RP are random unitary transforms, generalized compressed random precoding is a random compressed transform, and IBSFT-RP can be both a random unitary transform and a random compressed transform.

[0096] S2, the receiving antenna performs OFDM demodulation on the received signal and inputs it into a low-complexity cross-domain orthogonal approximate message passing receiver to construct a signal estimation problem containing random transformation prior information;

[0097] Specifically, the received signal for

[0098]

[0099] in is the transmitting antenna n t and receiving antenna n r The frequency domain channels between them. Define the row-column interleaving matrix and Rewrite the above formula as

[0100]

[0101] in in k=1,…,K, is the transmitting antenna n t and receiving antenna n r The kth subcarrier frequency response of the channel between is as follows: Figure 2 shown. is AWGN, the distribution remains unchanged. Therefore, the MIMO multipath channel can be decomposed independently by subcarrier, that is,

[0102]

[0103] in express express express

[0104] Therefore, signal detection in RP-MIMO-OFDM becomes an estimation problem with the following two constraints:

[0105]

[0106] Φ:s i ~P s ,

[0107] where k = 1, ..., K, the element s in s i Press P s Independent and identically distributed. When N rx =N tx = 1, A degenerates into a diagonal matrix. RP disturbs the symbol vector s, and the elements in s are no longer orthogonal on each subcarrier. Therefore, the low-complexity signal detection of RP-MIMO-OFDM can no longer use the original scheme. The following steps and systems of the present invention propose cross-domain orthogonal approximate message passing (CD-OAMP) for low-complexity signal detection. RP ensures that each symbol s i Transmit on all subchannels in A to achieve maximum diversity gain, thereby improving performance.

[0108] S3, the low-complexity cross-domain orthogonal approximate message passing receiver includes a linear estimator and a nonlinear estimator;

[0109] The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, inputs the output of the linear estimator into the nonlinear estimator for symbol-by-symbol MMSE demodulation after inverse random transformation, and outputs the output of the nonlinear estimator as a signal detection result after the iteration is completed.

[0110] Specifically, at the receiving end of the RP-MIMO-OFDM wireless communication system, the present invention designs a low-complexity cross-domain orthogonal approximate message passing (cross-domain OAMP, CD-OAMP) receiver, including a linear estimator (linearestimator, LE) and a nonlinear estimator (nonlinear estimator, NLE), and a random transformation module, such as Figure 3 shown.

[0111] CD-OAMP for unitary transformation RP

[0112] As mentioned above, the signal detection problem of RP-MIMO-OFDM is an estimation problem with the following two constraints:

[0113]

[0114] Φ:s i ~P s ,

[0115] where k = 1, ..., K, the element s in s i Press Ps Independent and identically distributed. LE and NLE are used to handle these two constraints respectively, and the random unitary transform and its inverse transform are responsible for converting the frequency domain and the random transform domain in the iteration.

[0116] (1)LE: Prior information from NLE and random transformation Assuming it as an independent and identically distributed (IID) Gaussian distribution, a linear minimum mean square error (LMMSE) estimation is performed under constraint Γ. For all subcarrier indices k, k = 1, ..., K, from t = 1 Start (t is the iteration index, are the pairs required for the tth iteration The estimated value and variance of , so the first iteration needs to set the initial value),

[0117]

[0118] in

[0119]

[0120] Since the frequency domain channel matrix A is block diagonal, N R ×N T dimensional LMMSE can be obtained by K times N rx ×N tx dimensional LMMSE implementation, thus greatly reducing the complexity.

[0121] (2) Random unitary transform and its inverse transform: After LE, the inverse random transform converts the Converted to random transform domain Similarly, the random transformation before LE will randomly transform s in the domain t Convert to frequency domain (Same as the random transform in the transmitter). For the random unitary transform and its inverse transform, the error variance Due to the Gaussian nature of the random matrix, after inverse random unitary transformation and orthogonalization, the input of NLE can be regarded as the Gaussian observation of s, that is,

[0122] (3)NLE: NLEφ t (·) Includes symbol-by-symbol MMSE demodulation and orthogonalization. In the random transform domain The output of s is estimated to be

[0123]

[0124] (4) Orthogonalization: The orthogonalization process of CD-OAMP is the same as that of OAMP, that is,

[0125]

[0126] Orthogonalization is an essential core operation of OAMP and is used to eliminate message dependencies during the message delivery process.

[0127] (5) Output: The final output of CD-OAMP is the MMSE output of the last iteration. Right now in is the number of iterations, usually a preset value, such as setting

[0128] (6) State evolution (SE) and optimality: The state evolution describes the dynamic performance of CD-OAMP by predicting the MSE of each iteration. SE is

[0129]

[0130] Among them, S~P s , The random precoding matrix Ξ can make the equivalent channel It has the right unitary invariant property, so SE can accurately describe the CD-OAMP dynamics of large-scale systems. T =KN tx Under the assumption that is large enough and SE has only one fixed point, CD-OAMP is the optimal replica MAP.

[0131] Inverse transformation problem in CD-OAMP of compression transformation RP

[0132] The main challenge facing CD-OAMP for compression transformation RP is that compression transformation is irreversible, and CD-OAMP still needs to solve the inverse transformation problem of compression transformation. Figure 4 As shown, the architecture can include generalized compressed random precoding. Specifically, the difficulty lies in compressing the internal permutation matrix Π in IBSFT l is the fat matrix of the incomplete permutation, in the compressed IBSFT transform Π l Some elements will be lost, which makes it irreversible. Therefore, in CD-OAMP, the inverse transformation should be The elements dropped from the LE input Make up for it in

[0133]

[0134] in For supplementary items, only The terms in are dropped, and the other terms are zero. After adding the dropped terms, the estimated variance becomes

[0135]

[0136] Specifically, the parameters used in this embodiment are as follows:

[0137] 1. Transmitter implementation

[0138] The following is a specific implementation of the 2×2 RP-MIMO-OFDM system, using MATLAB for software simulation. Figure 1 As shown, for a 2×2 MIMO system, N tx =N rx =2. The baseband parameters are set as follows: subcarrier spacing 15kHz, subcarrier number K = 400, and ideal bandwidth 6MHz. A MIMO-OFDM symbol frame vector contains 2K = 800 elements, that is, Random unitary transformation and random compression transformation are used for random precoding. The random unitary transformation adopts the FT-RP scheme, that is, Ξ=ΠF 2K , where Π is a 2K×2K dimensional random permutation matrix, F 2K is the normalized 2K-point FFT matrix; the random compression transform adopts the IBSFT-RP scheme, that is,

[0139]

[0140] in It is M s Point FFT matrix, Π l (l=1,…,L) is N s ×M s The inner permutation matrix, Π is the 2K×2K dimensional outer permutation matrix, where LN s =2K, compression ratio δ = M s / N s .

[0141] 2. MIMO multipath fading channel settings

[0142] The MIMO multipath fading channel is simulated using the TDL channel model. For the 2×2 RP-MIMO-OFDM system, there are 4 different multipath fading channels. The basic parameters of the channel are set as follows: the number of multipaths is 5, the power spectrum of these 5 paths is [10.18 0.17 0.15 0.12], the maximum delay is 1.5 microseconds, the transmit and receive filters use root raised cosine filters with a roll-off factor of β = 0.4, and the sampling rate is the same as the ideal bandwidth, that is, 6MHz. In the MATLAB simulation, the time domain impulse response of the channel is first randomly generated according to the power spectrum, expressed as a vector Then the frequency domain response diagonal matrix of the fading channel is obtained Thus, the entire MIMO multipath fading channel matrix is ​​obtained, that is,

[0143]

[0144] According to Figure 2 As shown, the MIMO multipath fading channel matrix consists of a diagonal matrix It can be transformed into a block diagonal matrix A.

[0145] In the simulation, after the transmitter and channel settings are completed, The received frequency domain signal can be obtained

[0146] 3. Receiver Implementation

[0147] The received frequency domain signal is detected using the CD-OAMP detector. Perform signal detection and set the maximum number of iterations to 6. The detection process of CD-OAMP is as follows:

[0148] make is the Lipschitz continuous symbol-by-symbol demodulation function, For the tth iteration The estimation function of

[0149]

[0150] From t = 1, x 1 =0 starts,

[0151]

[0152] in

[0153]

[0154] Independent of s, the final estimated output is Right now

[0155] 4. Complexity Analysis

[0156] The complexity of CD-OAMP of random unitary transformation RP is analyzed below, which mainly comes from three aspects: LE, NLE and random changes. For LMMSE in LE, the complexity of multiplication and matrix inversion in each iteration is or As for NLE, the complexity of symbol-by-symbol MMSE estimation is The complexity of random transformation in Haar-RP is In FT-RP In IBSFT-RP Therefore, in the RP-MIMO-OFDM system, for Haar-RP (completely random dense matrix), the time complexity of CD-OAMP is For FT-RP, the complexity is For IBSFT-RP, the complexity is in is the number of iterations. In contrast, due to the large-scale matrix inversion calculation, the time complexity of the original OAMP is as high as

[0157] FT-RP or IBSFT-RP has broad application prospects in small-scale RP-MIMO-OFDM systems with relatively low detection complexity. However, when the system scale is N tx As N increases, the complexity of the CD-OAMP detector increases. tx For large-scale RP-MIMO-OFDM systems, CD-OAMP becomes more complex and is no longer applicable. It should be replaced by CD-MAMP.

[0158] Corresponding to the embodiment of the aforementioned random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method, the present invention also provides a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication system, the system comprising:

[0159] A transmitter module and a receiver module;

[0160] The transmitter module includes a transmitting antenna and a precoding submodule; the symbol vector to be transmitted is randomly encoded by the precoding submodule, and the randomly encoded signal is transmitted to the receiver module using the antenna after OFDM modulation;

[0161] The receiver module includes a receiving end antenna and a low-complexity cross-domain orthogonal approximate message passing receiver; the receiving end antenna performs OFDM demodulation on the received signal and inputs it into the low-complexity cross-domain orthogonal approximate message passing receiver to construct a signal estimation problem containing random transformation prior information;

[0162] The low-complexity cross-domain orthogonal approximate message passing receiver includes a linear estimator, a nonlinear estimator and a random transformation module;

[0163] The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, inputs the output of the linear estimator into the nonlinear estimator for symbol-by-symbol MMSE demodulation after inverse random transformation, and outputs the output of the nonlinear estimator as a signal detection result after the iteration is completed.

[0164] Corresponding to the aforementioned embodiment of a random precoding MIMO OFDM wireless communication method, the present invention further provides an embodiment of a random precoding MIMO OFDM wireless communication device.

[0165] See also Figure 7 A random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication device provided in an embodiment of the present invention includes a memory and one or more processors, wherein the memory stores executable code, and when the processor executes the executable code, it is used to implement a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method in the above embodiment.

[0166] An embodiment of a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication device provided by the present invention can be applied to any device with data processing capabilities, and the any device with data processing capabilities can be a device or apparatus such as a computer. The device embodiment can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by the processor of any device with data processing capabilities in which it is located reading the corresponding computer program instructions in the non-volatile memory into the internal memory for execution. From a hardware perspective, if Figure 7 As shown, it is a hardware structure diagram of a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication device provided by the present invention, in which any device with data processing capability is located, except Figure 7 In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities in which the apparatus in the embodiments is located may also include other hardware, generally based on the actual functions of the device with data processing capabilities, which will not be described in detail.

[0167] The implementation process of the functions and effects of each unit in the above-mentioned device is specifically described in the implementation process of the corresponding steps in the above-mentioned method, and will not be repeated here.

[0168] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.

[0169] An embodiment of the present invention further provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method in the above embodiment is implemented.

[0170] The computer-readable storage medium may be an internal storage unit of any device with data processing capability described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium may also be an external storage device of any device with data processing capability, such as a plug-in hard disk, a smart media card (SMC), an SD card, a flash card, etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of any device with data processing capability. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and may also be used to temporarily store data that has been output or is to be output.

[0171] The present invention also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method is implemented.

[0172] Those skilled in the art will readily appreciate other embodiments of the present application after considering the description and practicing the contents disclosed herein. The present application is intended to cover any modification, use or adaptation of the present application, which follows the general principles of the present application and includes common knowledge or customary techniques in the art that are not disclosed in the present application. The description and examples are intended to be exemplary only, and the true scope and spirit of the present application are indicated by the claims.

[0173] It should be understood that the above general description and the detailed description below are only exemplary and explanatory and cannot limit the present application. The present application is not limited to the precise structure described above and shown in the drawings, and various modifications and changes can be made without departing from the scope thereof. The scope of the present application is limited only by the attached claims.

Claims

1. A random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method, characterized in that: The method includes: The transmitter performs random precoding on the symbol vector to be transmitted, and transmits the randomly coded signal to the receiver using the antenna after OFDM modulation; The receiving antenna performs OFDM demodulation on the received signal and inputs it into a low-complexity cross-domain orthogonal approximate message passing receiver to construct a signal estimation problem containing random transformation prior information; The low-complexity cross-domain orthogonal approximate message passing receiver includes a linear estimator and a nonlinear estimator; The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, inputs the output of the linear estimator into the nonlinear estimator for symbol-by-symbol MMSE demodulation after inverse random transformation, and outputs the output of the nonlinear estimator as a signal detection result after the iteration is completed.

2. A random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 1, characterized in that: The random precoding matrix construction method used in the random precoding includes random unitary transformation or random compression transformation.

3. The random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 2, characterized in that: The random unitary transformation or random compression transformation specifically includes: Random precoding based on Haar distribution unitary matrix: Perform singular value decomposition on an independent and identically distributed Gaussian matrix G, that is, G = U∑V H , where U and V are the unitary matrices of the Haar distribution; Random precoding based on fast transformation: random precoding matrix Ξ=ΠT N , where Π is a random permutation matrix, T N is the FT matrix, including the Fast Fourier Transform or Walsh-Hadamard Transform; Random precoding based on large-scale Gaussian matrix: For a matrix composed of independent and identically distributed Gaussian elements The N×N square matrix is ​​composed of. When N→∞, the square matrix is ​​regarded as a unitary matrix; Generalized Compressed Random Precoding: Using an incomplete shuffle matrix Π of size N×M (N<M) in Construct a generalized contraction random matrix, that is, Ξ=Π in U, where U is a random unitary matrix; Random precoding based on interleaved block sparse fast transform: The random precoding based on interleaved block sparse fast transform of size N×M is a random unitary transform when N = M and a random compression transform when N < M. It decomposes a vector into L blocks and interleaves multiple low-dimensional fast transform matrices, where N = LN s , M = LM s , that is, constructing a sparse matrix with random characteristics, namely in, It is M s Point fast transformation, Π l (l=1, ..., L) is N s ×M s The internal permutation matrix is ​​the N×N external permutation matrix; when Π l (l = 1, ..., L) is a positive matrix, that is, N s =M s When L=1, IBSFT is a random unitary transform, and when L=1, it degenerates into a random precoding based on fast transform.

4. The random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 1, characterized in that: The signal estimation problem of constructing the signal estimation problem including the random transformation prior information specifically includes: Received signal for in is the transmitting antenna n t and receiving antenna n r The frequency domain channel between and is a row-column interleaved matrix, rewriting the above formula as in in For AWGN, the distribution remains unchanged; The MIMO multipath channel is decomposed independently by subcarriers, that is, k=1,…,K, in express express express Signal detection is optimized as an estimation problem with the following two constraints: F:s i ~P s , where k = 1, ..., K, the element s in s i Press P s Independent and identically distributed. When N rx =N tx When =1, A degenerates into a diagonal matrix.

5. The random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 4, characterized in that: The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, and specifically includes: The prior information from NLE and random transformation Considered as independent and identically distributed Gaussian distribution, linear minimum mean square error estimation is performed under constraint Γ. For all subcarrier indices k, k = 1, ..., K, from t = 1, start, in 6. The random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 4, characterized in that: The inverse random transformation specifically includes: In the frequency domain Converted to random transform domain After inverse random unitary transformation and orthogonalization, the input of the nonlinear estimator is regarded as Gaussian observations of s, that is, 7. The random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 6, characterized in that: The nonlinear estimator performs symbol-by-symbol MMSE demodulation specifically as follows: Nonlinear estimator φ t (·) Includes symbol-by-symbol MMSE demodulation and orthogonalization, random transformation domain The output of s is estimated to be Orthogonalization:

8. The random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to claim 4, characterized in that: When precoding uses random compression transform, the inverse random transform needs to The elements dropped from the LE input Make up for it in in For supplementary items, only The items in are discarded, and the other items are zero.

9. A random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication system for implementing the method according to any one of claims 1 to 8, characterized in that: The system includes: A transmitter module and a receiver module; The transmitter module includes a transmitting antenna and a precoding submodule; the symbol vector to be transmitted is randomly encoded by the precoding submodule, and the randomly encoded signal is transmitted to the receiver module using the antenna after OFDM modulation; The receiver module includes a receiving end antenna and a low-complexity cross-domain orthogonal approximate message passing receiver; the receiving end antenna performs OFDM demodulation on the received signal and inputs it into the low-complexity cross-domain orthogonal approximate message passing receiver to construct a signal estimation problem containing random transformation prior information; The low-complexity cross-domain orthogonal approximate message passing receiver includes a linear estimator, a nonlinear estimator and a random transformation module; The linear estimator performs linear minimum mean square error estimation on the prior information from the nonlinear estimator and the random transformation, inputs the output of the linear estimator into the nonlinear estimator for symbol-by-symbol MMSE demodulation after inverse random transformation, and outputs the output of the nonlinear estimator as a signal detection result after the iteration is completed.

10. A random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication device, comprising a memory and one or more processors, wherein the memory stores executable code, characterized in that: When the processor executes the executable code, a random precoding multiple-input multiple-output orthogonal frequency division multiplexing wireless communication method according to any one of claims 1 to 8 is implemented.

Images