A high-parallelism and low-complexity equalization algorithm for wireless communication systems

Abstract

This invention discloses a highly parallel and low-complexity equalization algorithm for wireless communication systems, comprising: Step 1, setting an initial iteration value for the estimated vector s, and determining the initial iteration value s(0) of the estimated vector s based on the quadrant relationship between the elements in the estimated vector s and the elements in the vector after the transformation of the received vector y matrix; Step 2, calculating the estimated vector after iteration by designing an iterative strategy for the estimated vector s, calculating the fundamental matrix W, and setting the initial iteration value s(0) and iteration parameters, and obtaining the estimated vector s(T) after the Tth iteration through multiple iterations; Step 3, designing a constellation point search scheme, and calculating the final estimated value of the estimated vector s based on the estimated vector s(T) after the Tth iteration. The technical solution provided by this invention solves the problem that existing MIMO system equalization algorithms, which are based on traditional gradient search, suffer from low parallelism and high hardware implementation costs.

Description

Technical Field

[0001] This invention relates to, but is not limited to, the field of wireless communication technology, and particularly to a high-parallelism and low-complexity equalization algorithm for wireless communication systems. Background Technology

[0002] Industries such as information and communication technology (ICT), media, finance, and insurance are leading the current digital transformation. Key technologies supporting digitalization include software-defined devices, big data, blockchain, cybersecurity, virtual reality (VR), and augmented reality (AR). The increasing complexity of applications in daily life presents greater challenges for wireless communications.

[0003] Multiple-Input Multiple-Output (MIMO) technology, as a key technology in mobile communication, has attracted increasing attention due to its high network capacity, low latency, and high robustness. In massive MIMO systems, the number of antennas increases dramatically, leading to a sharp increase in computational complexity. The rapidly increasing number of antennas in massive MIMO systems results in highly complex and computationally intensive operations, posing a significant challenge to baseband processing algorithms and circuit implementation. Therefore, this paper proposes a technical requirement for practically applicable massive MIMO equalization algorithms.

[0004] Existing MIMO system equalization algorithms are based on traditional gradient search. However, these existing MIMO system equalization algorithms still face the following limitations in hardware implementation: First, traditional gradient search methods have low parallelism; second, traditional gradient search methods require multiple searches, resulting in high hardware implementation costs. Summary of the Invention

[0005] The purpose of this invention is to provide a highly parallel and low-complexity equalization algorithm for wireless communication systems. Existing MIMO system equalization algorithms, which are based on traditional gradient search, suffer from low parallelism and high hardware implementation costs.

[0006] The technical solution of this invention: This embodiment of the invention provides a high-parallelism and low-complexity equalization algorithm for wireless communication systems, comprising: estimating a transmitted signal s in a MIMO system based on the input-output relationship of the MIMO system using a high-parallelism and low-complexity equalization algorithm, wherein the transmitted signal s is the estimated vector s; the high-parallelism and low-complexity equalization algorithm includes:

[0007] Step 1, set the initial value of the estimation vector s for iteration, including: performing matrix transformation on the received vector y in the MIMO system, and determining the initial value s(0) of the estimation vector s based on the quadrant relationship between the elements in the estimation vector s and the elements in the vector after the matrix transformation of the received vector y.

[0008] Step 2: By designing an iterative strategy for estimating the vector s, the estimated vector after iteration is calculated, including: calculating the basic matrix W, setting the initial value s(0) and the iteration parameters, and obtaining the estimated vector s(T) after the Tth iteration through multiple iterations.

[0009] Step 3: Design a constellation point search scheme and calculate the final estimated value of the estimated vector s based on the estimated vector s(T) after the Tth iteration.

[0010] Optionally, in the high parallelism and low complexity equalization algorithm for wireless communication systems as described above, in the uplink of the MIMO system, Nr receiving antennas at the base station simultaneously communicate and transmit data with Nt users of a single antenna, and the input-output relationship of the MIMO system is expressed as: y = Hs + n.

[0011] Wherein, Nr >>;

[0012] The estimated vector s represents the transmitted signals of Nt users, and the size of the estimated vector s is Nt×1; This indicates the number of constellation points in a constellation chart;

[0013] The received vector y represents the signal received by Nr base station antennas, and the size of the received vector y is Nr×1;

[0014] H represents the Rayleigh fading channel matrix;

[0015] It is additive white Gaussian noise;

[0016] The high-parallelism and low-complexity equalization algorithm for wireless communication systems is used to estimate the transmitted signal s in y = Hs + n, i.e., s is the estimation vector.

[0017] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems as described above, in step 1,

[0018] Step 11, use the diagonally dominant matrix G = H H H performs a matrix transformation on the received vector y, and the matched filter vector y is obtained through the matrix transformation. MF =H H y has a size of Nt×1; where H is the channel matrix;

[0019] Step 12, based on the k-th element of the estimated vector s and the matched filter vector y obtained in step 11... MF =H H Given that the k elements of y lie in the same quadrant, the initial value of the estimated vector s, s(0), can be represented as:

[0020]

[0021] Step 13, based on y MF The quadrant containing the kth element determines the kth element of the initial value s(0) for iteration.

[0022] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems as described above, in step 13,

[0023] y MF When the k-th element is located in the first, second, third, and fourth quadrants respectively, the k-th element of the estimated vector s's initial value s(0) is set to...

[0024] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems described above, step 2 includes:

[0025] Step 21: Define the fundamental matrix W and calculate the fundamental matrix W;

[0026] Step 22, set the initial values ​​and initial parameters for iteration as: s(-1) = s(0), c(-1) = c(0) = y MF -Ws(0), α(-1)=α(0)=1, q(-1)=q(0)=Wc(0);

[0027] Step 23: Set the iteration number T, perform iterative calculation, and obtain the estimated value s(T) of the initially sent signal s after the Tth iteration.

[0028] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems described above,

[0029] In step 21, the basic matrix used as the intermediate calculation matrix according to the MMSE detection algorithm is set as: W = HHH + N0E s -1 I Nt ;

[0030] Where N0 is the noise variance, is the transmitted signal power, and I is an Nt×Nt identity matrix.

[0031] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems described above,

[0032] In step 22, the initial iteration value is set as: s(-1) = s(0); the initial parameters include: c(-1) = c(0) = y MF -Ws(0), α(-1)=α(0)=1, q(-1)=q(0)=Wc(0); where s, c and q are Nt×1 complex vectors, and α is a parameter;

[0033] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems described above,

[0034] In step 23, the following two calculations are performed in each iteration:

[0035] s(t+1)=α(t)(s(t)+β(t)c(t))+(1-α(t))s(t-1);

[0036] c(t+1)=α(t)(c(t)-β(t)q(t))+(1-α(t))c(t-1);

[0037] Furthermore, in each iteration, the three iteration parameters are calculated based on the results of the previous iteration:

[0038] q(t) = Wc(t);

[0039]

[0040]

[0041] After T iterations, s(T) is obtained, where s(T) represents the estimated value of the initially transmitted signal s after completing T iterations.

[0042] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems described above, step 3 includes:

[0043] The distance between the estimated vector s(T) after T iterations and all other constellation points in the constellation diagram is calculated using a constellation point search scheme, and the final estimated value of the estimated vector s is obtained based on the estimated vector s(T) after T iterations.

[0044] Optionally, in the high-parallelism and low-complexity equalization algorithm for wireless communication systems described above,

[0045] In step 3, the k-th element of the final estimated value of the estimated vector s is represented as:

[0046] s_k=2<0.5(s(T)_k+1))-1;

[0047] In this context, <*> represents rounding the asterisk (*) to the nearest whole number.

[0048] The beneficial effects of this invention: Based on the characteristics of large-scale MIMO systems, this invention provides a highly parallel and low-complexity equalization algorithm for wireless communication systems. It addresses the problems of low parallelism and high complexity in equalization algorithms for large-scale MIMO systems by proposing an improved highly parallel and low-complexity equalization algorithm. In the technical solution of this invention, based on the characteristics of large-scale MIMO systems, the specific calculation methods of the parallel iteration part, the quadrant-based initial value calculation part, and the constellation point search part can all reduce the complexity of the equalization algorithm and improve its accuracy. The technical solution provided by this invention specifically has the following beneficial effects:

[0049] (1) Reduce computational complexity: By using the balancing algorithm provided in this embodiment of the invention, the computational complexity of a large number of matrix inversions in the balancing algorithm can be reduced to O(Nt). 3 The complexity is reduced to O(Nt) 2 ).

[0050] (2) Improved parallelism: The equalization algorithm provided in this embodiment of the invention can avoid the large number of serial calculations in traditional equalization algorithms. Through iterative loops, each element in the estimated signal (i.e., the estimated vector s) can be calculated all at once, which greatly improves parallelism and is beneficial to the hardware implementation of the algorithm. Attached Figure Description

[0051] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0052] Figure 1 A schematic diagram of a MIMO system model used in the high-parallelism and low-complexity equalization algorithm for wireless communication systems provided in this embodiment of the invention.

[0053] Figure 2 for Figure 1 The illustrated embodiment provides a schematic diagram of the signal transmission and reception principle of the MIMO system.

[0054] Figure 3 The flowchart illustrates a high-parallelism and low-complexity equalization algorithm for wireless communication systems, provided as an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

[0056] As explained in the background section, in massive MIMO systems, the computational complexity increases dramatically with the number of antennas. The rapidly increasing number of antennas in massive MIMO systems leads to highly complex and computationally intensive operations, posing a significant challenge to baseband processing algorithms and circuit implementation. Therefore, a technical requirement for a practically applicable massive MIMO equalization algorithm is proposed.

[0057] However, existing MIMO system equalization algorithms, based on traditional gradient search, suffer from low parallelism and high hardware implementation costs. Therefore, there is an urgent need to find a practically applicable large-scale MIMO equalization algorithm that balances low complexity and high parallelism.

[0058] To address the issues of low parallelism and high hardware implementation cost in existing MIMO system equalization algorithms, this invention proposes an improved high-parallel equalization algorithm to solve the problems of low parallelism and high complexity in large-scale MIMO equalization algorithms. Specifically, it proposes a high-parallelism and low-complexity equalization algorithm for wireless communication systems.

[0059] The present invention provides the following specific embodiments, which can be combined with each other. For the same or similar concepts or processes, they may not be described again in some embodiments.

[0060] Figure 1 This is a schematic diagram of a MIMO system model used in a high-parallelism and low-complexity equalization algorithm for wireless communication systems, provided in an embodiment of the present invention. Figure 2 for Figure 1 The illustrated embodiment provides a schematic diagram of the signal transmission and reception principle of the MIMO system.

[0061] The high-parallelism and low-complexity equalization algorithm for wireless communication systems provided in this invention mainly includes the following three parts: a quadrant-based initial value calculation part, a parallel iteration part, and a constellation point search part.

[0062] like Figure 1 As shown, in the uplink of a MIMO system, the base station has Nr receiving antennas, which simultaneously communicate and transmit data with Nt users connected to a single antenna. The system model is as follows. Figure 1 As shown, the condition Nr >> is usually required. (See reference...) Figure 2 The signal transmission and reception principle shown is as follows: In this MIMO system, through channel coding, the bit stream information of Nt users is encoded in parallel and the results are mapped to points on the constellation diagram, finally generating signal vectors transmitted by Nt users.

[0063] Vector s∈Q, where vector s represents the transmitted signals of Nt users, and vector s has a size of Nt×1. Let y represent the set of points on the constellation diagram. Let vector y represent the signals received by Nr base station antennas, and the size of vector y is Nr×1. The input-output relationship of this MIMO system is shown in the following formula (1):

[0064] y = Hs + n; (1)

[0065] Formula (1) is the basic transmission model for signals in a MIMO system, where the transmitted signal is s and the received signal is y. The technical solution provided in this embodiment aims to estimate the transmitted signal s after the received signal y at the base station is known. In the above formula (1), H represents the Rayleigh fading channel matrix, where each element of the channel matrix H is an independent and identically distributed matrix with a mean of 0 and a variance of 1. It is additive white Gaussian noise, and each element of it follows an independent identically distributed system with a mean of 0 and a variance of σ². The main purpose of the equalization algorithm provided in this embodiment of the invention is to estimate the transmitted signal in the MIMO system, that is, to estimate the transmitted signal s in formula (1).

[0066] The high parallelism and low complexity balance algorithm for wireless communication systems provided in this embodiment of the invention mainly includes three parts: (1) setting initial values ​​for iteration; (2) designing iteration strategy; and (3) designing constellation point search scheme.

[0067] Figure 3 This is a flowchart illustrating a high-parallelism and low-complexity equalization algorithm for a wireless communication system, provided as an embodiment of the present invention. Figure 3 As shown, the high parallelism and low complexity equalization algorithm for wireless communication systems provided in this embodiment of the invention includes the following steps:

[0068] Step 1, set the initial values ​​for iteration:

[0069] Typically, the initial value for iteration is set to a zero vector because the prior information about the final solution is unknown. It's important to note that for the uplink of a large-scale MIMO system, when Nr >> , the channel matrix H is asymptotically orthogonal. Therefore, for an uplink large-scale MIMO system, the Gram matrix G = H. H H is diagonally dominant, and all elements on the diagonal are positive, with values ​​close to Nr. A matrix transformation is performed on the received vector y using the diagonally dominant matrix, resulting in the matched filter vector y. MF =H H Let y be an Nt×1 complex vector. This means that the k-th element of the estimated vector s is related to the matched filter vector y. MF =HH The k elements of y lie in the same quadrant; where the matched filter y MF It is an Nt×1 complex vector. Based on the above characteristics, the k-th element of the initial value s(0) can be determined by a point in the quadrant. Therefore, for Q-QAM modulation, the k-th element of the initial value s(0) of the estimated vector s can be expressed as the following formula (2):

[0070]

[0071] Therefore, based on the matched filter vector y MF The quadrant containing the k-th element determines the k-th element of the initial iteration value s(0). The matched filter vector y MF When the k-th element is located in the first, second, third, and fourth quadrants respectively, the k-th element of the estimated vector s's initial value s(0) is set to... After the initial value s(0) of the iteration is set, the iterative calculation is performed. The design scheme of the iterative measurement is described in detail in step two below.

[0072] Step 2, Design an iterative strategy:

[0073] First, the basic matrix W needs to be calculated. According to the MMSE detection algorithm, the basic matrix is ​​set as: W = H H H+N0E s -1 I Nt Where N0 is the noise variance, is the transmitted signal power, and I is an Nt×Nt identity matrix. Then, it is necessary to set the initial values ​​and initial parameters for the iteration, i.e., s(-1)=s(0), c(-1)=c(0)=y MF -Ws(0), α(-1)=α(0)=1, q(-1)=q(0)=Wc(0).

[0074] Where s, c, and q are three Nt×1 complex vectors, and α is a parameter. The fundamental matrix W is the matrix calculated during the intermediate process.

[0075] The iteration is performed T times, each time denoted by the letter t, i.e., t = 0, 1, ..., T-1. Each iteration performs the following two calculations:

[0076] s(t+1)=α(t)(s(t)+β(t)c(t))+(1-α(t))s(t-1); (3)

[0077] c(t+1)=α(t)(c(t)-β(t)q(t))+(1-α(t))c(t-1); (4)

[0078] The three parameters are calculated based on the results of the previous iteration, and their specific expressions are as follows:

[0079] q(t) = Wc(t); (5)

[0080]

[0081]

[0082] After T iterations, we can obtain s(T).

[0083] s(T) represents the estimated value of the initially sent signal s after completing T iterations.

[0084] Step 3, design a constellation point search scheme:

[0085] The constellation point search scheme can be used to calculate the distance between the estimated vector s(T) and all other constellation points in the constellation map, and finally obtain the final estimated value of the estimated vector s(T). Considering that s(T) is an Nt×1 vector, for Q-QAM, using existing equalization algorithms requires calculating the distance Nt×Q times. Therefore, the calculation of Euclidean distance will be simplified below. In the constellation map, the constellation points are arranged in a square, and the constellation map is divided into multiple square regions. The nearest point can be found using rounding. The k-th element of the final estimated value of the estimated vector s can be expressed as:

[0086] s_k=2(0.5(s(T)_k+1))-1; (8)

[0087] In this context, <*> represents rounding the asterisk (*) to the nearest whole number.

[0088] The present invention provides a highly parallel and low-complexity equalization algorithm for wireless communication systems. This algorithm addresses the problems of low parallelism and high complexity in equalization algorithms for large-scale MIMO systems by proposing an improved highly parallel and low-complexity equalization algorithm. In the technical solution of this invention, based on the characteristics of large-scale MIMO systems, the specific calculation methods of the quadrant-based initial value calculation part, the parallel iteration part, and the constellation point search part can all reduce the complexity of the equalization algorithm and improve its accuracy. The technical solution provided by this invention specifically has the following beneficial effects:

[0089] (1) Reduce computational complexity: By using the balancing algorithm provided in this embodiment of the invention, the computational complexity of a large number of matrix inversions in the balancing algorithm can be reduced to O(Nt). 3 The complexity is reduced to O(Nt) 2 );

[0090] (2) Improved parallelism: The equalization algorithm provided in this embodiment of the invention can avoid the large number of serial calculations in traditional equalization algorithms. Through iterative loops, each element in the estimated signal (i.e., the estimated vector s) can be calculated all at once, which greatly improves parallelism and is beneficial to the hardware implementation of the algorithm;

[0091] (3) The high parallelism and low complexity balance algorithm provided by this invention can be widely applied to the avionics systems of various types of fighter jets and civil aircraft that are currently in service and under development. It plays an important role in improving the communication efficiency of aircraft wireless communication systems, has broad market prospects, and has significant military, economic and social benefits.

[0092] The following is a specific implementation example illustrating the detailed implementation of the high parallelism and low complexity balancing algorithm for wireless communication systems provided in this invention.

[0093] Implementation Example

[0094] This implementation example uses a MIMO system with Nr=128, Nt=8, and 64-QAM modulation as an example for illustrative purposes.

[0095] The first step, based on the characteristics of the MIMO system, is to perform matrix transformation on the received vector y using a diagonally dominant matrix to calculate y. MF =H H y.

[0096] Wherein, the channel matrix H is a 128×8 complex matrix, the received vector y is a 128×1 complex vector, and the calculated result y after matrix transformation is... MF It is an 8×1 complex vector. Then, according to y... MF The initial value of the iteration is obtained by calculating the value of s(0). s(0) is an 8×1 complex vector, and the calculation method of its k-th element is determined by formula (8):

[0097] s(0))k=±4±4i; (9)

[0098] The second step is to calculate the 8×8 complex matrix W = H. H H+N0E s -1 I Nt And set s(-1) = s(0), c(-1) = c(0) = y MF -Ws(0), α(-1)=α(0)=1, q(--1)=q(0)=Wc(0).

[0099] Where s, c, and q are three 8×1 complex vectors, and α is a parameter. In this implementation example, to improve the performance of the MIMO system, the number of iterations is chosen to be 3, i.e., T = 3. Therefore, the value of t in each iteration is 0, 1, or 2.

[0100] When t=0, the iteration parameters are calculated as follows:

[0101] q(-1)=q(0)=Wc(0);

[0102]

[0103]

[0104] Subsequently, the estimated vector s and the iteration residual c are calculated as follows:

[0105] s(1)=α(0)(s(0)+β(o)c(0))+(1-α(0))s(--1);

[0106] c(1)=α(0)(c(0)-β(C)q(0))+(1-α(0))c(-1).

[0107] When t=1, the iteration parameters are calculated as follows:

[0108] q(1)=q(1)=Wc(1);

[0109]

[0110]

[0111] Subsequently, the estimated vector s and the iteration residual c are calculated as follows:

[0112] s(2)=α(1)(s(1)+β(1)c(1))+(1-α(1))s(0);

[0113] c(2)=α(1)(c(1)-β(1)q(1))+(1-α(1))c(0).

[0114] When t=2, the iteration parameters are calculated as follows:

[0115] q(2)=q(2)=Wc(2);

[0116]

[0117]

[0118] Subsequently, the estimated vector s and the iteration residual c are calculated as follows:

[0119] s(3)=α(2)(s(2)+β(2)c(2))+(1-α(2))s(1);

[0120] c(3)=α(2)(c(2)-β(2)q(2))+(1-α(2))c(1).

[0121] The third step is to calculate the final estimated value of the estimated vector s based on the value of the vector s(3) after three iterations for 64-QAM modulation, according to formula (8).

[0122] The high-parallelism and low-complexity equalization algorithm for wireless communication systems provided in this invention has the following advantages:

[0123] 1. Reduced complexity. Using the equilibrium algorithm of this invention, the complexity of performing numerous matrix inversions in the equilibrium algorithm can be reduced to O(Nt). 3 The complexity is reduced to O(Nt) 2 ).

[0124] 2. Improved Parallelism. The equalization algorithm of this invention avoids the extensive serial computation process of the original equalization algorithm. Through iterative iteration, each element of the estimated signal can be calculated all at once, significantly improving parallelism and facilitating the hardware implementation of the algorithm.

[0125] It should be noted that the high parallelism and low complexity equalization algorithm for wireless communication systems proposed in this invention can be widely applied to the avionics systems of various types of fighter jets and civil aircraft currently in service and under development. It plays an important role in improving the communication efficiency of aircraft wireless communication systems, has broad market prospects, and has significant military, economic and social benefits.

[0126] While the embodiments disclosed in this invention are as described above, they are merely illustrative of the embodiments to facilitate understanding of the invention and are not intended to limit the invention. Any person skilled in the art to which this invention pertains may make any modifications and variations in the form and details of the implementation without departing from the spirit and scope disclosed herein; however, the scope of patent protection for this invention shall still be determined by the scope defined in the appended claims.

Claims

1. A method for implementing a high-parallelism and low-complexity equalization algorithm for wireless communication systems, characterized in that, Based on the input-output relationship of the MIMO system, a high-parallelism and low-complexity equalization algorithm is used to estimate the transmitted signal s in the MIMO system, where the transmitted signal s is the estimated vector s. The high-parallelism and low-complexity equalization algorithm includes: Step 1, set the initial value of the estimation vector s for iteration, including: performing matrix transformation on the received vector y in the MIMO system, and determining the initial value s(0) of the estimation vector s based on the quadrant relationship between the elements in the estimation vector s and the elements in the vector after the matrix transformation of the received vector y. Step 2: By designing an iterative strategy for estimating the vector s, the estimated vector after iteration is calculated, including: calculating the basic matrix W, setting the initial value s(0) and the iteration parameters, and obtaining the estimated vector s(T) after the Tth iteration through multiple iterations. Step 3: Design a constellation point search scheme and calculate the final estimated value of the estimated vector s based on the estimated vector s(T) after the Tth iteration. In step 1, Step 11, use the diagonally dominant matrix G = H H H performs a matrix transformation on the received vector y, and the matched filter vector y is obtained through the matrix transformation. MF =H H y has a size of Nt×1; where H is the channel matrix; Step 12, based on the k-th element of the estimated vector s and the matched filter vector y obtained in step 11... MF =H H Given that the k elements of y lie in the same quadrant, the k-th element of the initial value s(0) of the estimated vector s is represented as: in, This indicates the number of constellation points in a constellation chart; Step 13, based on y MF The quadrant containing the kth element determines the kth element of the initial iteration value s(0); Step 2 includes: Step 21: Define the fundamental matrix W and calculate the fundamental matrix W; Step 22, set the initial values ​​and initial parameters for iteration as: s(-1) = s(0), c(-1) = c(0) = y MF -Ws(0), α(-1)=α(0)=1, q(-1)=q(0)=Wc(0); Step 23: Set the number of iterations T, perform iterative calculations, and obtain the estimated value s(T) of the initially sent signal s after the Tth iteration; In step 23, the following two calculations are performed in each iteration: s(t+1)=α(t)(s(t)+β(t)c(t))+(1-α(t))s(t-1); c(t+1)=α(t)(c(t)-β(t)q(t))+(1-α(t))c(t-1); Furthermore, in each iteration, the three iteration parameters are calculated based on the results of the previous iteration: q(t) = Wc(t); After T iterations, s(T) is obtained, where s(T) represents the estimated value of the initially transmitted signal s after T iterations. Step 3 includes: The distance between the estimated vector s(T) after T iterations and all other constellation points in the constellation diagram is calculated using a constellation point search scheme, and the final estimated value of the estimated vector s is obtained based on the estimated vector s(T) after T iterations.

2. The method for implementing a high-parallelism and low-complexity equalization algorithm for a wireless communication system according to claim 1, characterized in that, In the uplink of the MIMO system, Nr receiving antennas at the base station simultaneously communicate and transmit data with Nt users using a single antenna. The input-output relationship of the MIMO system is expressed as: y = Hs + n. Where Nr >> Nt; The estimated vector s represents the transmitted signals of Nt users, and the size of the estimated vector s is Nt×1; The received vector y represents the signal received by Nr base station antennas, and the size of the received vector y is Nr×1; H represents the Rayleigh fading channel matrix; It is additive white Gaussian noise; The method described above for implementing a high-parallelism and low-complexity equalization algorithm for a wireless communication system is used to estimate the transmitted signal s in y = Hs + n, i.e., s is the estimation vector.

3. The method for implementing a high-parallelism and low-complexity equalization algorithm for a wireless communication system according to claim 1, characterized in that, In step 13 y MF When the k-th element is located in the first, second, third, and fourth quadrants respectively, the k-th element of the estimated vector s's initial value s(0) is set to...

4. The method for implementing a high-parallelism and low-complexity equalization algorithm for a wireless communication system according to claim 1, characterized in that, In step 21, the basic matrix used as the intermediate calculation matrix is ​​set as follows according to the MMSE detection algorithm: W = H H H+N0E s - 1 I Nt ; Where N0 is the noise variance, is the transmitted signal power, and I is an Nt×Nt identity matrix.

5. The method for implementing a high-parallelism and low-complexity equalization algorithm for a wireless communication system according to claim 4, characterized in that, In step 22, the initial iteration value is set as: s(-1) = s(0); the initial parameters include: c(-1) = c(0) = y MF -Ws(0), α(-1)=α(0)=1, q(-1)=q(0)=Wc(0); where s, c and q are Nt×1 complex vectors, and α is a parameter.

6. The method for implementing a high-parallelism and low-complexity equalization algorithm for a wireless communication system according to claim 1, characterized in that, In step 3, the k-th element of the final estimated value of the estimated vector s is represented as: s_k = 2 < 0.5(s(T)_k+1) > -1; In this context, <*> represents rounding the asterisk (*) to the nearest whole number.

Images