CN106487723B - Channel estimation method and apparatus suitable for single-antenna interference cancellation technology - Google Patents

Abstract

The present invention provides a channel estimation method suitable for single-antenna interference cancellation technology, which includes obtaining a signal sequence estimation value by pre-equalizing a signal; restoring the sequence estimation value to the form of a burst sequence, and determining a direction by signal energy at the same time. and performing adaptive sequence estimation on the half-edge sequence containing the training sequence according to the obtained reference value of the burst sequence and the direction decision. The method of the present invention includes three parts: preprocessing, sequence restoration and direction determination, and adaptive channel estimation. The invention also provides a channel estimation device suitable for the single-antenna interference cancellation technology, which can be adapted to the single-antenna interference cancellation technology of the decoupling linear filtering nonlinear equalization method, and improves the performance of asynchronous interference cancellation.

Description

Channel estimation method and apparatus suitable for single-antenna interference cancellation technology

technical field

The present invention relates to the field of wireless communication technologies, and in particular, to a channel estimation method and device suitable for single-antenna interference cancellation technology.

Background technique

Co-channel interference (CCI, co-channel interference) is caused by multiple users transmitting information in the same channel. In the GSM system of adjacent cell communication, co-channel interference must be considered as a major interference. According to the condition of the terminal in the downlink, the single-antenna interference cancellation technology (SAIC, single-antenna-interference-cancellation) is undoubtedly a better method to solve the problem of co-channel interference suppression.

Conventional SAIC methods include MIC (Mono Interference Cancellation), joint detection, decoupling linear filter nonlinear equalization and other methods. MIC has insufficient degrees of freedom and harsh operating conditions, especially it cannot adapt to 8PSK modulated interference signals. Joint detection suppresses interference Strong capability, but high complexity, so considering the performance and complexity, the decoupling linear filter nonlinear equalization method can be used to suppress interference. Since the method does not need to estimate the channel condition of the interfering signal, the complexity can be reduced. At the same time, because the method is insensitive to the modulation type, it can be applied to both GMSK and 8PSK modulated signals, and the application area is wider than that of MIC.

From the principle of decoupling linear filtering nonlinear equalization technology, to improve the interference suppression performance of this method, we can start from three aspects: one is to improve the filter coefficient generation method to improve the filtering performance; The third is to provide a longer reference sequence based on the original training sequence. Among them, the latter two are to obtain more effective filter coefficients under the fixed filter coefficient generation method. Due to the existence of asynchronous interference, if the channel impulse response is simply estimated by the training sequence, the reliability will decrease, and interference may be introduced to cause the result to deviate from the expected channel estimation value. Close to the expected channel value, improve the ability of decoupling linear filter to suppress asynchronous interference.

SUMMARY OF THE INVENTION

In order to solve the problems existing in the prior art, the present invention provides a channel estimation method suitable for single-antenna interference cancellation technology, including:

The estimated value of the signal sequence is obtained by pre-equalizing the signal;

restoring the sequence estimate to the form of a burst sequence, and determining an adaptive channel estimation direction by the burst sequence; and

According to the obtained reference value of the burst sequence and the determined adaptive estimation direction, perform adaptive sequence estimation on the half-side sequence including the training sequence, and when it is determined that the adaptive estimation channel direction is the transition band, only the training sequence segment is Do adaptive channel estimation.

Optionally, in the process of determining the adaptive channel estimation direction, the original burst is divided into half-sequences with a length of 78 symbols on the left and right sides from the middle point, the average power values of the two sides are calculated respectively, and the average power ratio of the left and right sides is obtained. The adaptive channel estimation direction is determined as the direction with the smaller average power.

Optionally, after the adaptive channel estimation direction is selected, the adaptive channel estimation is performed from the side of the training sequence that is different from the adaptive direction, and after the adaptive channel estimation is completed, the adaptive channel estimation is performed from the side different from the adaptive channel estimation. The iteration starts on one side of the adaptive channel estimation direction.

Optionally, according to the magnitude of the average power ratio of the half-sequence on both sides, a ratio range belonging to the transition band is introduced, and within this range, it is determined that only the adaptive channel estimation of the training sequence segment is performed.

Optionally, before obtaining the signal sequence estimation value, obtain the first channel estimation value of the signal; in the adaptive sequence estimation, update the step size of the adaptive channel estimation according to the determined adaptive channel estimation direction , and the first channel estimation value is used as the initial value of the adaptive channel estimation.

Optionally, use the LS channel estimation method to obtain the first channel estimation value of the signal.

Optionally, when it is determined that the adaptive channel estimation direction is the transition band, adaptive estimation is only performed on the training sequence segment, and iterations are performed from the left side of the training sequence segment to the rightmost end of the training sequence.

Optionally, when performing the adaptive channel estimation, the LMS adaptive channel estimation method is adopted, and the linear average value of gradients of 2-3 points around the detection point is selected as the gradient value of the selected point.

Optionally, the method for restoring the estimated value of the signal sequence to the form of burst is to add head and tail bits, add a training sequence in the middle, and then perform GMSK or 8PSK modulation according to the modulation form of the desired signal to obtain the burst sequence. Reference.

A channel estimation device suitable for single-antenna interference cancellation technology, comprising:

a signal pre-equalization device for performing signal pre-equalization to obtain a signal sequence estimation value;

means for restoring the sequence estimate to the form of a burst sequence;

means for calculating the corresponding average power value of the burst sequence to determine the direction of adaptive channel estimation; and

A device for performing adaptive sequence estimation on a half-side sequence including a training sequence according to the reference value of the burst sequence and the adaptive channel estimation direction, and when it is determined that the adaptive estimated channel direction is a transition band, only the training sequence segment is adapted channel estimation.

Optionally, in the process of determining the adaptive channel estimation direction, the original burst is divided into half-sequences with a length of 78 symbols on the left and right sides from the middle point, the average power values on both sides are calculated respectively, and the average power ratio on the left and right sides is obtained. Determine The adaptive channel estimation direction is the direction with the smaller average power.

Optionally, after the adaptive channel estimation direction is determined, adaptive channel estimation is performed from the side of the training sequence that is different from the adaptive direction, and after the adaptive channel estimation is completed, the adaptive channel estimation is performed from the side different from the adaptive channel estimation. One side of the adaptive orientation starts to iterate.

Optionally, according to the magnitude of the average power ratio of the half-sequence on both sides, a ratio range belonging to the transition band is introduced, and within this range, it is determined that only the adaptive channel estimation of the training sequence segment is performed.

Optionally, before obtaining the signal sequence estimation value, obtain the first channel estimation value of the signal; in the adaptive channel estimation, update the step size of the adaptive channel estimation according to the determined adaptive channel estimation direction, and The first channel estimation value is used as the initial value of the adaptive channel estimation.

Optionally, use the LS channel estimation method to obtain the first channel estimation value of the signal.

Optionally, when it is determined that the adaptive estimation channel direction is the transition band, adaptive estimation is only performed on the training sequence segment, and the iteration is performed from the left side of the training sequence segment to the rightmost end of the training sequence.

Optionally, when performing the adaptive channel estimation, the LMS adaptive channel estimation method is adopted, and the linear average value of gradients of 2-3 points around the detection point is selected as the gradient value of the selected point.

Optionally, the method for restoring the estimated value of the signal sequence to a burst sequence is to add head and tail bits, add a training sequence in the middle, and then perform GMSK or 8PSK modulation according to the modulation format of the desired signal to obtain the burst sequence. Reference.

The present invention provides a channel estimation method suitable for single-antenna interference cancellation technology, which includes obtaining a signal sequence estimation value through pre-equalization of a signal; restoring the sequence estimation value to the form of a burst sequence, and at the same time obtaining a signal sequence estimation value through pre-equalization; The power value determines a direction decision; and according to the obtained reference value of the burst sequence and the direction decision, adaptive channel estimation is performed on the half-edge sequence containing the training sequence. The method of the invention includes three parts: preprocessing, sequence restoration, direction determination, and adaptive channel estimation, which can be adapted to the single-antenna interference elimination technology of the decoupling linear filtering nonlinear equalization method, and improve the performance of asynchronous interference elimination.

Description of drawings

FIG. 1 is a schematic structural diagram of the decoupling linear filtering nonlinear equalization SAIC according to an embodiment of the present invention;

FIG. 2 is a flowchart of a channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention;

FIG. 3 is a flowchart of adaptive channel estimation direction selection in the channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention;

4A to 4D respectively correspond to the channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention, in which the desired signal is interfered by the synchronous co-channel, part of the data segment of the desired signal is not interfered, half of the desired signal is covered by interference, and the desired signal training sequence segment Undisturbed 4 typical cases;

5 is a BER curve of a desired signal subjected to synchronous co-channel interference in the channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention;

FIG. 6 is a BER curve of the uninterrupted BER curve of the desired signal part data segment in the channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention;

7 is a BER curve in which half of the desired signal is covered by interference in the channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention;

FIG. 8 is a BER curve in which the desired signal training sequence segment is not interfered in the channel estimation method of the single-antenna interference cancellation technology according to an embodiment of the present invention.

Detailed ways

The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become apparent from the following description and claims. It should be noted that, the accompanying drawings are all in a very simplified form and use imprecise ratios, and are only used to facilitate and clearly assist the purpose of explaining the embodiments of the present invention.

The channel estimation device of the single antenna interference cancellation technology (SAIC) suitable for the decoupling linear filter nonlinear equalization proposed by the present invention is located in the decoupling linear filter nonlinear equalization SAIC system as shown in 102 in Fig. 1. In step 101, after double sampling, it is divided into two channels of single data bits to perform channel estimation, respectively, to provide channel estimation values for CCI filtering in step 103. Step 102, that is, the specific implementation method of channel estimation is as follows:

First set the standard form of the received signal r as follows

where r(n) is the nth symbol of the received signal, h is the desired channel impulse response, a is the desired sequence, g is the impulse response of the ith interfering user, b is the sequence of the ith user, and n (n) is Gaussian white noise.

Then, the corresponding modules are operated according to the flow shown in FIG. 2 .

1. Pre-estimate the bits received in step 201 in step 202 to obtain a first channel estimation value. This step channel estimation adopts the most basic LS channel estimation. In order to improve the accuracy of the channel estimation value, 26-point correlation estimation can be used instead of 16-point correlation estimation to obtain the channel estimation value.

h _LS = (A ^T A) ^-1 Ar (1-2)

2. In step 203, the channel pre-estimation value h _LS obtained in step 202 and the received sequence r obtained in step 201 are pre-equalized. As for the equalization method, MLSE is used for the GMSK signal, and RSSE or DFSE is used for the 8PSK signal to obtain the desired value. The first estimate a of the sequence.

3. In step 204, the expected burst form is restored from the obtained first estimated value a of the expected sequence. The specific method is to add the head and tail bits, add the training sequence in the middle, and then perform GMSK or 8PSK modulation according to the modulation form of the desired signal to obtain the burst estimated value for subsequent reference.

4. In step 205, make an adaptive channel estimation direction judgment according to the received sequence. Divide the Burst of length 156 into left and right parts from the middle, each with 78 points, calculate the average power value of both sides respectively, and obtain the average power ratio of the left and right sides.

Since there is no need to know the specific boundary position of the interference, there is no need to use an algorithm to precisely locate the interference, but only to determine the direction by judging the average power ratio of the left and right sides. The criterion takes a limit n (n>1) as a reference point, and is divided into three cases: p≤1/n, 1/n<p<n, p≥n. The first is to do adaptive channel estimation to the left, and the third is There are three types of adaptive channel estimation to the right, the second is the transition band, and the adaptive range is limited to the training sequence range. When n is too close to 1, too sensitive direction judgment will make the adaptive channel estimation difficult to converge, and if n is too large, it is not conducive to deal with the asynchronous interference situation.

为自适应参考序列值，以步骤205的方向判断值为自适应迭代遍历方向。选择方向的自适应信道估计具体流程按图3所示：5. Next, in step 206, perform LMS adaptive channel estimation, use the h _LS value obtained in step 202 as the initial value of channel estimation, and use the sequence in step 204 to restore the value

is the adaptive reference sequence value, and the direction judgment value in step 205 is the adaptive iterative traversal direction. The specific flow of the adaptive channel estimation of the selected direction is shown in Figure 3:

(1) According to the judgment value of the adaptive iterative traversal direction in step 205, determine whether it is in the transition zone, if so, go to step 303, otherwise, go to step 302.

从训练序列的最左侧向训练序列最右侧做自适应信道估计，迭代式为(2) Step 303 is executed at this time, the range of the adaptive channel estimation is limited to the training sequence segment, the step size of the adaptive channel estimation is updated in step 303, and the step 304 is to take the leftmost group of the training sequence segment and the corresponding

Adaptive channel estimation is performed from the far left of the training sequence to the far right of the training sequence, and the iterative formula is

where μ is the step size, which determines the convergence rate and the final error. The size of μ, the theoretical optimum is

Among them, λ _max is the maximum eigenvalue of the autocorrelation function matrix _RX of the training sequence segment of the received sequence r, and λ _min is the minimum eigenvalue of the autocorrelation function matrix _RX of the received sequence r. The actual value is adjusted based on the test results.

ε(n) is expressed as

is the estimated error value.

(3) If it is not a transition zone, then step 302 is executed to further determine whether to perform the adaptive traversal process to the left or right. Taking the left as an example, step 305 is performed. First, the adaptive step size μ is updated, and the algorithm is the same as (1-5). Then, step 306 is performed, and a group whose rightmost end of the training sequence is taken as the starting point is taken as the starting point.

从训练序列的最右侧向burst序列最左侧做自适应信道估计，迭代式同(1-4)。如果判定方向为向右，则按步骤307和步骤308，从训练序列的最左侧向burst序列的最右侧做自适应信道估计。and the corresponding

Adaptive channel estimation is performed from the far right of the training sequence to the far left of the burst sequence, and the iterative formula is the same as (1-4). If it is determined that the direction is to the right, according to steps 307 and 308, adaptive channel estimation is performed from the leftmost part of the training sequence to the rightmost part of the burst sequence.

进行平滑操作。平滑操作采用线性平均值的形式，即取一段值的梯度的均值来代替原来某个点的梯度值。设当前自适应点的梯度值为Nk，则取长度为L段内的梯度值得均值来代替原梯度值。于是，实际梯度值变为(4) During the traversal process, in order to improve the convergence of the adaptive channel estimation and reduce the influence of the interference boundary and burst interference on the adaptive channel estimation, in steps 304, 306 and 308, the gradient in the adaptive channel estimation is value

Smooth operation. The smoothing operation takes the form of a linear average, that is, the average of the gradients of a range of values is taken to replace the original gradient value of a certain point. Suppose the gradient value of the current adaptive point is Nk, then take the mean value of the gradient value in the L segment to replace the original gradient value. So, the actual gradient value becomes

否则，则继续从起始点开始，再做一次遍历。After each traversal, it is judged whether the maximum number of iterations has been reached, if so, it ends, and the updated channel impulse response value is output.

Otherwise, continue from the starting point and do another traversal.

Figures 5 to 8 show the comparison charts of the BER curves in three cases of TU50 channel, SNR=25dB, no SAIC and adaptive channel estimation processing, SAIC processing, SAIC+ adaptive channel estimation processing. Figures 5 to 8 are the interfering signal delays of 0, 30%, 50%, and 70%, respectively, as shown in Figures 4A to D, which means that the desired signal is synchronously interfered, part of the desired signal data is not interfered, and half of the desired signal is interrupted. There are four typical cases: interference and expected signal training sequence segment without interference. When the interference is close to synchronization or the interference does not cover the training sequence segment, the LMS adaptive estimation results are basically the same. The reason is that the estimation is performed directly in the training sequence segment at this time, and the result is closer to the ideal estimated value, while when the interference is offset , the adaptive channel estimation utilizes the information of the uncovered interference segment and utilizes more reference points (except the training sequence), which can improve the quality of the channel estimation and thus the performance. It shows that the present invention can show relatively good gain when dealing with asynchronous co-channel interference.

The present invention also provides a channel estimation device suitable for the single-antenna interference cancellation technology, including:

a signal pre-equalization device for performing signal pre-equalization to obtain a signal sequence estimation value;

means for restoring the sequence estimate to a burst sequence;

means for calculating the left and right average power values of the burst sequence to determine the direction of the adaptive channel estimation; and

A device for performing adaptive channel estimation on the half-side sequence including the training sequence according to the reference value of the burst sequence and the determined direction of the adaptive channel estimation, and when the direction of the adaptively estimated channel is determined to be the transition zone, only the training The sequence segment is used for adaptive channel estimation.

The present invention provides a channel estimation method suitable for single-antenna interference cancellation technology, which includes obtaining a signal sequence estimation value through pre-equalization of a signal; restoring the sequence estimation value to the form of a burst sequence, and at the same time obtaining a signal sequence estimation value through pre-equalization; The power value determines the direction of the adaptive channel estimation; and according to the obtained reference value of the burst sequence and the determined direction of the adaptive channel estimation, the adaptive channel estimation is performed on the half-side sequence including the training sequence. The method of the invention includes three parts: preprocessing, sequence restoration, direction determination, and adaptive channel estimation, which can be adapted to the single-antenna interference elimination technology of the decoupling linear filtering nonlinear equalization method, and improve the performance of asynchronous interference elimination.

Obviously, those skilled in the art can make various changes and modifications to the invention without departing from the spirit and scope of the invention. Thus, provided that these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.

Claims (12)

1. A channel estimation method applicable to single-antenna interference cancellation technology, characterized in that, comprising: The estimated value of the signal sequence is obtained by pre-equalizing the signal; The estimated sequence value is restored to the form of a burst sequence, and the adaptive channel estimation direction is determined by the burst sequence, wherein the method of restoring the estimated sequence value to the form of the burst sequence is to add head and tail bits , add the training sequence in the middle, and then perform GMSK or 8PSK modulation according to the modulation form of the desired signal to obtain the reference value of the burst sequence. The length of the burst sequence is 156 symbols. In the process of determining the direction of adaptive channel estimation , the burst sequence is divided into half-sequences with equal left and right lengths and 78 symbols from the middle point, the average power values of the left and right sides are calculated respectively, and the average power ratio p of the left and right sides is obtained. A limit n greater than 1 is the reference point , when p≤1/n, determine the direction of adaptive channel estimation to the left, when p≥n, determine the direction of adaptive channel estimation to the right, when 1/n<p<n, determine the direction of adaptive channel estimation the direction is the transition zone; and According to the obtained reference value of the burst sequence and the determined adaptive channel estimation direction, the adaptive channel estimation is performed on the half-side sequence including the training sequence, and when the adaptive channel estimation direction is determined to be the transition band, only the training sequence segment for adaptive channel estimation. 2. The channel estimation method applicable to the single-antenna interference cancellation technology according to claim 1, wherein after the adaptive channel estimation direction is determined, the method from the training sequence that is different from the adaptive channel The adaptive channel estimation starts from one side of the estimation direction, and after each adaptive channel estimation is completed, the adaptive channel estimation is performed again from the side different from the adaptive channel estimation direction for iteration until the maximum iteration is reached. frequency. 3. The channel estimation method applicable to the single-antenna interference cancellation technology according to claim 1, wherein, before obtaining the signal sequence estimation value, the first channel estimation value of the signal is obtained; in the adaptive channel estimation , the step size of the adaptive channel estimation is updated according to the determined adaptive channel estimation direction, and the first channel estimation value is used as the initial value of the adaptive channel estimation. 4. The channel estimation method applicable to the single-antenna interference cancellation technology according to claim 3, wherein the first channel estimation value of the signal is obtained by using the LS channel estimation method. 5. The channel estimation method suitable for single-antenna interference cancellation technology as claimed in claim 1, wherein when it is determined that the adaptive channel estimation direction is a transition band, adaptive channel estimation is only performed on the training sequence, and the adaptive channel estimation is performed from the training sequence. The sequence starts at the far left of the sequence and iterates until the far right of the training sequence. 6. The channel estimation method applicable to the single-antenna interference cancellation technology according to claim 1, wherein, when the adaptive channel estimation is performed, the LMS adaptive channel estimation method is adopted, and the selected detection point is 2 - The linear mean of the gradients of 3 points as the gradient value of the selected point. 7. A channel estimation device suitable for single-antenna interference cancellation technology, characterized in that it comprises: a signal pre-equalization device for performing signal pre-equalization to obtain a signal sequence estimation value; A device for restoring the estimated sequence value to the form of a burst sequence, wherein the method for restoring the estimated sequence value to the form of the burst sequence is to add head and tail bits, add the training sequence in the middle, and then according to The modulation form of the desired signal is subjected to GMSK or 8PSK modulation to obtain the reference value of the burst sequence; A device for calculating the corresponding average power value of the burst sequence to determine an adaptive channel estimation direction, wherein the burst sequence has a length of 156 symbols, and in the process of determining the adaptive channel estimation direction, the burst sequence From the middle point, it is divided into half-sequences of equal length on the left and right and both of them are 78 symbols, calculate the average power values of the left and right sides respectively, and obtain the average power ratio p of the left and right sides. Taking a limit n greater than 1 as the reference point, when p≤1 When /n, the adaptive channel estimation direction is determined to the left; when p≥n, the adaptive channel estimation direction is determined to the right; when 1/n<p<n, the adaptive channel estimation direction is determined to be the transition zone; as well as A device for performing adaptive channel estimation on a half-side sequence including a training sequence according to the reference value of the burst sequence and the determined adaptive channel estimation direction, and when it is determined that the adaptive channel estimation direction is a transition band, Only do adaptive channel estimation for the training sequence segment. 8. The channel estimation device suitable for the single-antenna interference cancellation technique according to claim 7, wherein after the adaptive channel estimation direction is determined, from the training sequence different from the adaptive channel estimation The adaptive channel estimation starts from one side of the direction, and after each adaptive channel estimation is completed, the adaptive channel estimation is performed again from the side different from the adaptive channel estimation direction for iteration until the maximum number of iterations is reached. . 9. The channel estimation device suitable for single-antenna interference cancellation technology according to claim 7, characterized in that, before obtaining the signal sequence estimation value, the first channel estimation value of the signal is obtained; , the step size of the adaptive channel estimation is updated according to the determined adaptive channel estimation direction, and the first channel estimation value is used as the initial value of the adaptive channel estimation. 10 . The channel estimation device suitable for single-antenna interference cancellation technology according to claim 9 , wherein the first channel estimation value of the signal is obtained by using the LS channel estimation method. 11 . 11. The channel estimation device suitable for single-antenna interference cancellation technology according to claim 7, wherein when it is determined that the adaptive channel estimation direction is a transition band, adaptive channel estimation is only performed on the training sequence, and the The sequence starts at the far left of the sequence and iterates until the far right of the training sequence. 12. The channel estimation device suitable for single-antenna interference cancellation technology according to claim 7, characterized in that, when the adaptive channel estimation is performed, the LMS adaptive channel estimation method is adopted, and the selected detection point is 2 points before and after the detection point. - The linear mean of the gradients of 3 points as the gradient value of the selected point.

Images