阅读说明：本技术 一种用于短波通信的Turbo时域均衡方法 (turbo time domain equalization method for short wave communication ) 是由 张健 张凯 邱利利 王小军 牛磊 冯永乾 王鹏 阎旭 于 2019-09-26 设计创作，主要内容包括：本发明公开了一种用于短波通信的Turbo时域均衡方法,首先,使用MMSE均衡器对接收信号进行均衡；再将其输入到译码器进行译码；而后利用译码器输出的软信息来构造平均信息和方差信息；利用方差外信息矩阵计算第j时刻的横向滤波器抽头系数；结合已知的多径信道参数和平均外信息来模拟接收信号,从而得到信的估计；最后,MMSE均衡器利用模拟接收信号r和真实接收信号r再次对多径信号进行均衡。本发明避免了时域算法中复杂的矩阵求逆运算,有效降低了迭代均衡的处理复杂度,在不降低均衡性能的前提下,既能同时消除线性和非线性干扰,又能大大降低计算复杂度。(The invention discloses a Turbo time domain equalization method for short wave communication, which comprises the following steps of firstly, equalizing a received signal by using an MMSE equalizer; then inputting the data into a decoder for decoding; then, soft information output by the decoder is utilized to construct average information and variance information; calculating a tap coefficient of a transverse filter at the j time by using the variance external information matrix; simulating the received signal by combining the known multipath channel parameters and the average extrinsic information, thereby obtaining an estimate of the signal; finally, the MMSE equalizer equalizes the multipath signal again by using the analog received signal r and the real received signal r. The invention avoids complex matrix inversion operation in a time domain algorithm, effectively reduces the processing complexity of iterative equalization, can simultaneously eliminate linear and nonlinear interference and greatly reduce the calculation complexity on the premise of not reducing the equalization performance.)

1. A Turbo time domain equalization method for short wave communication is characterized by comprising the following steps:

Step 1, setting short wave communication channel total p '+ q' +1 order, wherein, there is p 'order before the main path, there is q' order after the main path, and the multipath channel parameter h (t) ═ h_-p′(t)，…，h₀(t)，…，h_q′(t)), the user information bit u to be transmitted is channel coded and then BPSK modulated, and the mapping symbol x ═ (x)) is obtained₀，x₁，…，x_n-1) The transmitting end transmits data symbols x^prefixafter the data symbol is transmitted through a multipath channel, a receiving end receives a signal r;

Wherein the channel coded codeword bit is c ═ (c)₀，c₁，…，c_i，…，c_n-1)，c_i∈{0，1}；x_i＝1-2×c_iN is the codeword length;

Step 2, the receiving end adopts a low-complexity Turbo time domain equalization method based on MMSE equalizationEqualizing and decoding multipath fading of received signal, and outputting estimated code word

2. The Turbo time-domain equalization method for short-wave communication of claim 1, wherein in step 1, the data symbol x is^prefix＝(x_-l，…，x_-1，x₀，x₁，…，x_j，…，x_n-1，x_n，…，x_n+l-1)；

Wherein (x)_-l，…，x_-1) As a front guard band, (x)_n，…，x_n+l-1) A rear protective belt; the front guard band and the rear guard band are respectively a symbol sequence composed of { -1, +1} and have lengths respectively equal to or greater than p '+ q'.

3. The Turbo time-domain equalization method for short wave communication according to claim 1, wherein the mathematical expression of the received value of the received signal r at the t-th time is as follows:

r_t＝h_-p′(t)x_p′+t+…+h_-1(t)x_1+t+h₀(t)x_t+h₁(t)x_-1+t+…+h_q′(t)x_-q′+t+n_t，

(-p′＜＝t＜n+q′)

Wherein n is_tFor the t-th moment, the obedient mean value is 0 and the variance isThe two-dimensional normal distribution additive white Gaussian noise sampling value is obtained;

4. The Turbo time domain equalization method for short wave communication according to claim 1, wherein the receiving end equalizes and decodes the multipath fading of the received signal by using a low-complexity Turbo time domain equalization method based on MMSE equalization, which specifically comprises:Approximating an inverse of the channel characteristic with a transversal filter of finite order; i.e. the equaliser uses a priori information L provided by the decoder^out(b) equalizing the multipath fading signal, extracting the soft information of the j bit, traversing all j to obtain the soft information of all symbols/bits, and further carrying out next decoding; the method is implemented according to the following steps:

(1) Initialization: let the order of the transversal filter be p + q +1, the maximum number of iterations be L, the current number of iterations L be 0, the initial soft information is set to be neutral information, i.e. L^out(b)＝(0，…，0，…，0)；

(2) Judging whether the current iteration frequency meets L < L, if so, turning to the step 3; otherwise, jumping to step 6;

(3) making the bit serial number j of the soft information equal to 0;

(4) judging whether the bit sequence number of the soft information meets j < n, if so, turning to (4.1), otherwise, skipping to the step 5;

(4.1) constructing a channel circulation matrix H_j；

(4.2) constructing an average extrinsic information vectorAnd variance extrinsic information matrix

Wherein the content of the first and second substances,(-q′-q≤j′≤p′+p)；

namely, it is

Wherein the content of the first and second substances,diag denotes the main diagonal;

respectively orderthen the average extrinsic information vector can be obtainedSum variance extrinsic information matrixThe specific expression is as follows:

(4.3) calculating a tap coefficient w (j) of the transversal filter:

wherein w (j) ═ w_q(j)，…，w₀(j)，…，w_-p(j))^Tthe superscript T is the transposition operation, I_p+qRepresenting a unit matrix of order p + q (.)^HRepresenting a conjugate transpose operation, (.)^-1Represents the inversion operation, C represents the q' + q +1 th column of matrix H;

(4.4) calculating the output of the transversal Filterwherein r is_j＝(r_j-q，…，r_j，…，r_j+p)^T；

Calculating the mean value mu_jSum variance

(4.5) calculating the soft information of the j bit

(4.6) adding 1 to the variable j, and jumping to the step 4;

(5) Soft information L output by equalizerⁱⁿ(b) sending the signal to a decoder for decoding, and outputting L by the decoder^out(b)；

(6) To L^out(b is 0) making hard decision to obtain estimated code wordif it isif the code word constraint is met, exiting; otherwise, adding 1 to l; and judging whether the iteration times meet L ═ L, if so, exiting, otherwise, jumping to the step 2.

5. The Turbo time-domain equalization method for short wave communication of claim 4, wherein the constructed channel cyclic matrix H_jThe method specifically comprises the following steps:

Constructing a channel cyclic matrix H at the time t_t：

Wherein H_tHas dimensions p + q +1 rows and p + q + p '+ q' +1 columns.

6. The Turbo time domain for short wave communications of claim 4equalization method, characterized in that, setting a channel matrix as a constant value within a limited number of symbol periods, said channel circulant matrix H is then used_jChannel circulant matrix H at time t_tThe method is simplified as follows:

wherein H_tEach of the 2 nd to p + q +1 th rows of (a) is a result of the cyclic right shift of the previous row of the row, respectively.

7. the Turbo time-domain equalization method for short-wave communication of claim 4, wherein the decoder outputs L^out(b) Comprises the following steps:

L^out(b)＝(L^out(b₀)，L^out(b₁)，…，L^out(b_n-1))；

Wherein, the symbol L^out(b_i) Represents the log-likelihood ratio of the ith bit of the decoder output, andIf L is^out(b_i) If the number of bits is more than or equal to 0, judging the ith bit as 0; otherwise, judging as 1; p (b)_i0) is the probability that the ith bit is determined to be 0.

Technical Field

The invention belongs to the technical field of short-wave communication, and particularly relates to a Turbo time domain equalization method for short-wave communication.

background

For a wireless channel, multipath propagation, fast-changing frequency selectivity and time selective fading are main factors causing distortion of a transmission signal, inter-symbol interference is introduced into channel impulse response, so that a received signal is seriously damaged, and an equalizer is generally required to be used in a receiver to eliminate the influence of inter-symbol interference and then decode the received signal. Turbo equalization is a technology for carrying out data detection by repeatedly iterating equalization and decoding, and by the characteristics of random coding and iterative decoding, excellent performance close to the Shannon limit is obtained, so that the Turbo equalization becomes an equalizer mainly adopted in short-wave communication at present.

In a classical Turbo equalization algorithm, a channel equalizer performs soft information calculation by using a Maximum A Posteriori (MAP) probability criterion, the MAP equalizer is an optimal algorithm for minimizing a bit error rate of a symbol, and the main purpose is to provide channel receiving soft information for a decision decoder, that is, to calculate a possible information symbol probability under a condition of a known received symbol. However, the algorithm involves a large number of logarithm and multiplication operations, the operation complexity increases exponentially with the memory length of the channel and the modulation order, and the operation is complex and not beneficial to engineering implementation.

Disclosure of Invention

In order to solve the above problems, the present invention aims to provide a Turbo time domain equalization method for short wave communication, which adopts a low complexity Turbo equalization method based on Minimum Mean Square Error (MMSE) criterion, avoids complex matrix inversion operation in a time domain algorithm, effectively reduces processing complexity of iterative equalization, and can simultaneously eliminate linear and nonlinear interference and greatly reduce calculation complexity without reducing equalization performance.

In order to achieve the above object, the present invention adopts the following technical solutions.

A Turbo time domain equalization method for short wave communication comprises the following steps:

Step 1, setting short wave communication channel total p '+ q' +1 order, wherein, there is p 'order before the main path, there is q' order after the main path, and the multipath channel parameter h (t) ═ h_-p′(t)，…，h₀(t)，…，h_q′(t)), the user information bit u to be transmitted is channel coded and then BPSK modulated, and the mapping symbol x ═ (x)) is obtained₀，x₁，…，x_n-1) The transmitting end transmits data symbols x^prefixAfter the data symbol is transmitted through a multipath channel, a receiving end receives a signal r;

wherein the channel coded codeword bit is c ═ (c)₀，c₁，…，c_i，…，c_n-1)，c_i∈{0，1}；x_i＝1-2×c_iN is the codeword length;

Step 2, the receiving end adopts a low-complexity Turbo time domain equalization method based on MMSE equalization to equalize and decode the multipath fading of the received signal, and outputs an estimated code word

Further, the data symbol x^prefix＝(x_-1，…，x_-1，x₀，x₁，…，x_j，…，x_n-1，x_n，…，x_n+l-1)；

wherein (x)_-l，…，x_-1) As a front guard band, (x)_n，…，x_n+l-1) A rear protective belt; the front guard band and the rear guard band are respectively a symbol sequence composed of { -1, +1} and have lengths respectively equal to or greater than p '+ q'.

Further, the mathematical expression of the received value of the received signal r at the t-th time is as follows:

r_t＝h_-p′(t)x_p′+t+…+h_-1(t)x_1+t+h₀(t)x_t+h₁(t)x_-1+t+…+h_q′(t)x_-q′+t+n_t，

(-p′＜＝t＜n+q′)

Wherein n is_tFor the t-th moment, the obedient mean value is 0 and the variance isThe two-dimensional normal distribution additive white Gaussian noise sampling value is obtained;

Further, the receiving end uses a low-complexity Turbo time domain equalization method based on MMSE equalization to equalize and decode the multipath fading of the received signal, which specifically comprises: approximating an inverse of the channel characteristic with a transversal filter of finite order; i.e. the equaliser uses a priori information L provided by the decoder^out(b) equalizing the multipath fading signal, extracting the soft information of the j bit, traversing all j to obtain the soft information of all symbols/bits, and further carrying out next decoding; the method is implemented according to the following steps:

(1) Initialization: let the order of the transversal filter be p + q +1, the maximum number of iterations be L, the current number of iterations L be 0, the initial soft information is set to be neutral information, i.e. L^out(b)＝(0，…，0，…，0)；

(2) Judging whether the current iteration frequency meets L < L, if so, turning to the step 3; otherwise, jumping to step 6;

(3) Making the bit serial number j of the soft information equal to 0;

(4) judging whether the bit sequence number of the soft information meets j < n, if so, turning to (4.1), otherwise, skipping to the step 5;

(4.1) constructing a channel circulation matrix H_j；

(4.2) constructing an average extrinsic information vectorAnd variance extrinsic information matrix

wherein the content of the first and second substances,(-q′-q≤j′≤p′+p)；

Namely, it is

Wherein the content of the first and second substances,diag denotes the main diagonal;

Respectively orderThen the average extrinsic information vector can be obtainedsum variance extrinsic information matrixThe specific expression is as follows:

(4.3) calculating a tap coefficient w (j) of the transversal filter:

Wherein w (j) ═ w_q(j)，…，w₀(j)，…，w_-p(j))^TThe superscript T is the transposition operation, I_p+qRepresentation orderunit array of p + q (.)^HRepresenting a conjugate transpose operation, (.)^-1Represents the inversion operation, C represents the q' + q +1 th column of matrix H;

(4.4) calculating the output of the transversal Filterwherein r is_j＝(r_j-q，…，r_j，…，r_j+p)^T；

Calculating the mean value mu_jSum variance

(4.5) calculating the soft information of the j bit

(4.6) adding 1 to the variable j, and jumping to the step 4;

(5) Soft information L output by equalizerⁱⁿ(b) Sending the signal to a decoder for decoding, and outputting L by the decoder^out(b)；

(6) To L^out(b is 0) making hard decision to obtain estimated code wordIf it isIf the code word constraint is met, exiting; otherwise, adding 1 to l; and judging whether the iteration times meet L ═ L, if so, exiting, otherwise, jumping to the step 2.

further, the channel circulation matrix H is constructed_jThe method specifically comprises the following steps:

Constructing a channel cyclic matrix H at the time t_t：

Wherein H_tthe dimension of (a) is p + q +1 rows and p + q + p '+ q' +1 columns;

further, setting the channel matrix to be a constant value within a limited number of symbol periods, then the channel cyclic matrix H at time t_tthe method is simplified as follows:

Wherein H_teach of the 2 nd to p + q +1 th rows of (a) is a result of the cyclic right shift of the previous row of the row, respectively.

Further, the decoder outputs L^out(b) Comprises the following steps:

L^out(b)＝(L^out(h₀)，L^out(b₁)，…，L^out(b_n-1))；

wherein, the symbol L^out(b_i) Represents the log-likelihood ratio of the ith bit of the decoder output, andif L is^out(b_i) If the number of bits is more than or equal to 0, judging the ith bit as 0; otherwise, judging as 1; p (b)_i0) is the probability that the ith bit is determined to be 0.

compared with the prior art, the invention has the beneficial effects that: the low-complexity Turbo equalization algorithm based on MMSE equalization is adopted, the complex matrix inversion operation in a time domain algorithm is avoided, the processing complexity of iterative equalization is effectively reduced, linear and nonlinear interference can be eliminated simultaneously on the premise of not reducing the equalization performance, and the calculation complexity can be greatly reduced.

Drawings

The invention is described in further detail below with reference to the figures and specific embodiments.

FIG. 1 is a block diagram of an implementation flow of the present invention;

FIG. 2 is a block diagram of a multipath time-varying channel of the present invention;

FIG. 3 is a block diagram of a transversal filter in an embodiment of the invention;

FIG. 4 is a diagram of the time-varying characteristic of a short-wave channel with fading parameters of 2ms/1Hz in an embodiment of the present invention;

FIG. 5 is a diagram of a simulation result of a Turbo equalization algorithm and a corresponding simplified algorithm in an embodiment of the present invention;

FIG. 6 is a graph illustrating the impact of iteration count on performance of a simplified Turbo equalization algorithm in an embodiment of the present invention.

Detailed Description

the embodiments and effects of the present invention will be described in further detail below with reference to the accompanying drawings.

Referring to fig. 1, a Turbo time domain equalization method for short wave communication performs equalization in an iterative manner. Firstly, equalizing a received signal by using an MMSE equalizer; secondly, inputting the data into a decoder for decoding; then, the decoder uses the output soft information L^out(b) To construct average informationSum variance informationUsing an extrinsic information matrix of varianceCalculating a transverse filter tap coefficient at the j time; combining known multipath channel parameters and averaged extrinsic informationTo simulate received signalsThereby obtaining an estimate of the signal; finally, the MMSE equalizer utilizes the analog received signalAnd the true received signal r again equalizes the multipath signal.

It should be noted that, in fig. 1, the multipath channel parameters are usually obtained by a "channel estimation" module, and the accuracy of estimation directly affects the performance of the Turbo equalization system. For simplicity, the low complexity Turbo equalization technique is described in detail below, assuming that the receiving end has full knowledge of the channel parameters.

The method specifically comprises the following steps:

Step 1, setting short wave communication channel total p '+ q' +1 order, wherein, there is p 'order before the main path, there is q' order after the main path, and the multipath channel parameter h (t) ═ h_-p′(t)，…，h₀(t)，…，h_q′(t)), the user information bit u to be transmitted is channel coded and then BPSK modulated, and the mapping symbol x ═ (x)) is obtained₀，x₁，…，x_n-1) The transmitting end transmits data symbols x^prefixafter the data symbol is transmitted through a multipath channel, a receiving end receives a signal r; wherein the channel coded codeword bit is c ═ (c)₀，c₁，…，c_i，…，c_n-1)，c_i∈{0，1}；x_i＝1-2×c_iN is the codeword length;

Establishing an information sending model:

suppose that the channel has p '+ q' +1 order, the main path is preceded by p 'order and followed by q' order, and the characteristic at the t-th time is h (t) ═ h_-p′(t)，…，h₀(t)，…，h_q′(t)), a logic diagram is shown in fig. 2. In view of the complexity of the actual communication environment, the communication system must have multipath resistance. Therefore, guard bands of length l are added before and after the transmission of information, i.e.

x^prefix＝(x_-l，…，x_-1，x₀，x₁，…，x_j，…，x_n-1，x_n，…，x_n+l-1)，

Wherein (x)_-l，…，x_-1) As a front guard band, (x)_n，…，x_n+l-1) A rear guard band. The front guard band and the back guard band can be any sequence consisting of { -1, +1} and the lengths of the front guard band and the back guard band can be different, but the lengths of the front guard band and the back guard band are both equal to or greater than p '+ q'. Here, for convenience, front and back are referred toThe guard bands are the same in length and are all l (l is more than or equal to p + q). From the above parameters, the mathematical expression of the received value at the t-th time is

r_t＝h_-p′(t)x_p′+t+…+h_-1(t)x_1+t+h₀(t)x_t+h₁(t)x_-1+t+…+h_q′(t)x_-q′+t+n_t，

(-p′＜＝t＜n+q′)

Wherein n is_tFor the t-th moment, the obedient mean value is 0 and the variance isThe two-dimensional normally distributed Additive White Gaussian Noise (AWGN) sample value;

Then, after the transmitted data symbol is degraded through multipath channel transmission, the receiving end receives the signal r.

Step 2, the receiving end adopts a low-complexity Turbo time domain equalization method based on MMSE equalization to equalize and decode the multipath fading of the received signal, and outputs an estimated code word

A low complexity Turbo equalization structure is shown in figure 3. The essence is that the filter of 'infinite' order is used to approximate the inverse of the channel characteristic, thus achieving the purpose of counteracting the multipath channel effect. In practical applications, due to the limitations of hardware conditions and implementation complexity, a finite order is usually adopted to implement the transversal filter. For example, in fig. 3, an order of p + q +1 is used to approximate an inverse channel having a (positive) channel order of p '+ q' + 1. In the figure z^-1representing a unit time delay. The output of the transversal filter is the pair symbol x_jIs estimated.

the equalizer uses a priori information L provided by the decoder^out(b) And (3) equalizing the multipath fading signals, extracting the soft information of the j bit, traversing all j, and obtaining the soft information of all symbols/bits, so that a new decoding process is started, the error rate is reduced by one step, and the reliability of the system is improved.

The method is implemented according to the following steps:

(1) Initialization: let the order of the transversal filter be p + q +1, the maximum number of iterations be L, the current number of iterations L be 0, the initial soft information is set to be neutral information, i.e. L^out(b)＝(0，…，0，…，0)；

(2) Judging whether the current iteration frequency meets L < L, if so, turning to the step 3; otherwise, jumping to step 6;

(3) Making the bit serial number j of the soft information equal to 0;

(4) judging whether the bit sequence number of the soft information meets j < n, if so, turning to (4.1), otherwise, skipping to the step 5;

(4.1) constructing a channel circulation matrix H_j；

(4.2) constructing an average extrinsic information vectorand variance extrinsic information matrix

wherein the content of the first and second substances,(-q′-q≤j′≤p′+p)；is a size reflecting the decoder pair symbolwhen the estimation is very accurate, there is Indicates the degree of deviation (dispersion) of the evaluation results ifLarger indicates greater dispersion.

namely, it is

Wherein the content of the first and second substances,diag denotes the main diagonal;

Respectively orderThen the average extrinsic information vector can be obtainedSum variance extrinsic information matrixThe specific expression is as follows:

(4.3) calculating a tap coefficient w (j) of the transversal filter:

Wherein, I_p+qRepresenting a unit matrix of order p + q (.)^HRepresenting a conjugate transpose operation, (.)^-1Represents the inversion operation, C represents the q' + q +1 th column of matrix H;

(4.4) calculating the output of the transversal FilterWherein r is_j＝(r_j-q，…，r_j，…，r_j+p)^T；

Estimate for transversal filterCan be regarded as a random variable, the mean value mu of which_jSum varianceare respectively as

variance (variance)Can be viewed as the jitter introduced by the transversal filter and can be equated to a noise superimposed on the mean with a noise variance of

(4.5) calculating the soft information of the j bit

(4.6) adding 1 to the variable j, and jumping to the step 4;

(5) Soft information L output by equalizerⁱⁿ(b) Sending the signal to a decoder for decoding, and outputting L by the decoder^out(b)；

L^out(b)＝(L^out(b₀)，L^out(b₁)，…，L^out(b_n-1))；

Wherein, the symbol L^out(b_i) Represents the log-likelihood ratio of the ith bit of the decoder output, andIf L is^out(b_i) If the number of bits is more than or equal to 0, judging the ith bit as 0; otherwise, judging as 1; p (b)_i0) is the probability that the ith bit is determined to be 0.

(6) To L^out(b is 0) making hard decision to obtain estimated code wordIf it isif the code word constraint is met, exiting; otherwise, adding 1 to l; and judging whether the iteration times meet L ═ L, if so, exiting, otherwise, jumping to the step 2.

In the above process, the equalizer operates based on the information provided by the decoder, so that the bit soft information vector L output by the decoder^out(b) Referred to as a priori information.

In calculating the symbol x_jWithout using the symbol x output from the decoder for the estimation of (2)_jAny of (3). For example, the information vector will be averaged both when calculating the tap coefficients w (j) and the output of the transversal filterSum variance information matrixCorresponding toAndSet to 0 and 1, respectively. Hence the scalefor averaging extrinsic information vectors, scaleis an out of variance information matrix.

In one embodiment of the invention, the channel circulant matrix H_jThe method specifically comprises the following steps:

constructing a channel cyclic matrix H at the time t_t：

wherein H_tHas dimensions p + q +1 rows and p + q + p '+ q' +1 columns.

In another embodiment of the present invention, the time-varying channel has a time-varying characteristic that not only fades the signal but also produces a doppler spectrum. As shown in fig. 4, a characteristic diagram of a two-path fading parameter of a 2ms/1Hz short wave channel (real part) is shown. Two paths are parameters, 2ms is a time interval between the two paths, 1Hz is a doppler spectrum bandwidth of 1Hz caused by relative motion of the transceiver and multi-angle reflection, and it can be simply understood that the larger the value is, the faster the channel change is.

Under such time-varying channel, the channel circulation matrix H_tDifferent at each time, even if each row is different at the same time. This results in a huge calculation amount due to the inverse operation of the matrix when calculating the tap coefficients of the transversal filter, which is very disadvantageous for the engineering implementation. However, since the channel variation is not so severe over a limited number of symbol periods, the approximation can be regarded as a constant value, on the basis of which H can be assumed to be_tThe method is simplified as follows:

Wherein H_tEach of the 2 nd to p + q +1 th rows of (a) is a result of the cyclic right shift of the previous row of the row, respectively.

simulation experiment

In order to verify the performance of the method, the LDPC code is used as an error correcting code to perform simulation under the baseband condition; the signal-to-noise ratio in simulation is defined asthe unit is dB; e_sIn order for the received energy of the symbol to be, Is a gaussian white noise variance superimposed on the channel.

The simulation parameters are set as follows:

LDPC code: selecting LDPC codes (516, 1032) adopted by a mobile broadband wireless access standard (IEEE 802.16e), wherein the code rate is 0.5, the base matrix is 12 multiplied by 24, and the expansion factor is 43; the LDPC decoding Algorithm is Sum-Product decoding Algorithm (SPA), and the iteration number is 50;

multipath channel: the short wave time-varying channel has double path of 2ms/1 Hz.

It should be noted that, in the simulation, the performance of the Turbo equalization algorithm is mainly examined, so that it is assumed that the receiving end knows the parameters of the multipath channel, and in practical application, the channel estimation algorithm may be adopted to obtain the parameters related to the multipath channel.

simulation 1:

Inspecting the performance of a Turbo time domain equalization algorithm and a corresponding simplified algorithm under a short-wave time-varying channel; the number of joint iterations is 3, wherein the order of the transversal filter in the MMSE equalization algorithm is 20, and the simulation performance is shown in fig. 5.

As can be seen from fig. 5, both the Turbo time domain equalization algorithm and the corresponding simplified algorithm can effectively overcome the multipath effect in the multipath environment, and it can be seen that the corresponding simplified algorithm and the non-simplified algorithm have nearly the same performance. For example, when the bit error rate BER is 10^-5both algorithms have a desired signal-to-noise ratio of 12dB, with little performance difference between the two.

Simulation 2:

The iteration times directly influence the overall performance of the algorithm, and the performance is better as the iteration times are larger. The simulation inspects the influence of the combined iteration times of the simplified Turbo equalization algorithm on the performance under a short-wave time-varying channel, wherein the combined iteration times are respectively set to be 1 and 3. The simulation results are shown in fig. 6. It can be seen from the graph that the performance of iteration 3 is significantly better than that of iteration 1, and there is a performance difference of about 0.3-0.4dB between the two algorithms.

Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.

the above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.