A low complexity peak cancellation method for nr

Abstract

本发明公开了一种用于NR的低复杂度消峰方法，包括：将输入信号x输入0.5延迟的数字滤波器，获得x0p5；求x的幅值xF、x的相角xA、x0p5的幅值x0p5F和x0p5的相角x0p5A；系统带宽配置RB小于设定的带宽阈值时，对噪声做一个滑动平均滤波；根据幅值和相角求出该峰值点的复数值；对xflt做滑窗检测；xn_in(idx)中若ps(idx)＝0则通过低通滤波器filter0压缩带宽，若ps(idx)＝1则通过低通滤波器filter1压缩带宽；计算获得xc(idx)；将xc(idx)通过一个0.25延迟的全通滤波器。本发明在一倍采样率完成消峰功能，通过对峰值做滑动滤波器处理，在输入信号的带宽很小时，对峰值做一个滑动平均滤波处理，使得其消峰效果很好。

Description

Technical Field

[0001] This invention relates to a low-complexity peak removal method for NR, belonging to the field of low-complexity peak removal technology. Background Technology

[0002] The most commonly used clipping method is oversampling iteration plus filtering. However, oversampling iteration plus filtering has high implementation complexity, and it is difficult to design filters for small bandwidth and high sampling rate, resulting in poor peak reduction effect. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to overcome the defects of the prior art and provide a low-complexity peak removal method for NR.

[0004] To achieve the above objectives, this invention provides a low-complexity peak removal method for NR (Noise Reduction), comprising:

[0005] Step 4: When the system bandwidth configuration RB is less than the set bandwidth threshold, perform a moving average filter on the noise.

[0006]

[0007] In the formula, x flt The signal is after being filtered by moving average, where k is the moving average window length, and x... n_flt The input noise amplitude is idx, and idx is the index.

[0008] Prior to step four, the following steps are performed:

[0009] Step 5: For x flt Perform a sliding window test:

[0010]

[0011] p pos =[]

[0012] p pos =p pos ∪idx;

[0013] In the formula, P pos It records the peak value. The noise value is the merged value, and L is the sliding window detection length.

[0014] Prior to step four, perform the following steps:

[0015] Step 1: Input the input signal x into a digital filter with a 0.5 delay to obtain x. 0p5 .

[0016] Prior to step four, perform the following steps:

[0017] Step 2, calculate the amplitude of x. F The phase angle of x is x A x 0p5 amplitude x 0p5F and x 0p5 phase angle x 0p5A :

[0018]

[0019] In the formula, |·| represents finding the modulus of a complex number; ang(·) represents finding the phase angle function of a complex number.

[0020] Prior to step two, the following steps are performed:

[0021] Step 3: Obtain the noise value.

[0022]

[0023]

[0024] And record the corresponding index idx, A m This is the peak elimination threshold;

[0025] If x no or x 0p5no If it is greater than 0, then compare x. no and x 0p5no size:

[0026]

[0027]

[0028]

[0029] x can be calculated using the above formula. n_flt (idx), p s (idx) and x ang (idx).

[0030] Prioritize, in step three, if x F or x 0p5F If the amplitude does not exceed the set threshold, then x will be... no or x 0p5no The noise value is set to 0.

[0031] Prior to step five, the following steps are performed:

[0032] Step Six: Calculate the complex value of the peak point based on the amplitude and phase angle:

[0033]

[0034] In the formula, 1i is the imaginary part.

[0035] Prior to step six, the following steps are performed:

[0036] Step 7, x n_in If p in (idx) s If (idx) = 0, the bandwidth is compressed by using the low-pass filter filter0.

[0037] If p s If (idx) = 1, the bandwidth is compressed by using the low-pass filter filter1.

[0038] If p s (idx) = 0 and idx ∈ p pos ,but

[0039] x ot (k,idx)=x n_in (k)×C0(j), j=0,...,M-1

[0040] k = idx + j

[0041] In the formula, x ot (.) represents the value obtained by multiplying the complex value of the noise by the value of the filter coefficient, M is the coefficient length of the filter, and C0(.) is the value of the filter0 coefficient;

[0042] If p s (idx) = 1 and idx ∈ p pos ,but

[0043] x ot (k,idx)=x n_in (k)×C1(j), j=0,...,M-1

[0044] k = idx + j

[0045]

[0046] In the formula, C1(.) is the coefficient value of filter1, and x0(.) is the sum of the complex values ​​of noise at different positions multiplied by the filter coefficient values;

[0047] Step 8: Subtract x0(idx) from the input signal x(idx):

[0048] x c (idx) = x(idx) - x0(idx)

[0049] Using the above formula, x can be calculated. c (idx).

[0050] Prioritize step nine by repeating steps one through eight until n times, where n is a positive integer, to obtain the final x. c (idx);

[0051] Step 10, x c (idx) passes through an all-pass filter with a delay of 0.25.

[0052] Step 11, then repeat steps 1 through 8 until n times.

[0053] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0054] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0055] The beneficial effects achieved by this invention are as follows:

[0056] This invention achieves peak reduction at a sampling rate of 1. By performing a sliding filter on the peak values, when the bandwidth of the input signal is very small, a sliding average filter is applied to the peak values, resulting in a very good peak reduction effect. Attached Figure Description

[0057] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0058] The following embodiments are only used to illustrate the technical solutions of the present invention more clearly, and should not be used to limit the scope of protection of the present invention.

[0059] Step 1: Pass the input signal x through a digital filter with a 0.5-sampling delay to achieve a 0.5-sampling delay in the input signal, denoted as 0.5 phase. Let x be the input signal. 0p5 The input signal that does not pass through the digital filter with a delay of 0.5 is denoted as phase 0.

[0060] The 0.5-delay digital filter is an all-pass digital filter with an even number of taps.

[0061] Step 2: Calculate x separately 0p5 amplitude x F and phase angle x A x 0p5 amplitude x 0p5F and phase angle x 0p5A :

[0062]

[0063] In the formula, |·| represents finding the modulus of a complex number; ang(·) represents finding the phase angle of a complex number;

[0064] Step 3: Judge the amplitude of phase 0 and phase 0.5 respectively, calculate the sum of points that exceed the amplitude threshold (the threshold is set according to the peak-to-average power ratio of the input peak suppression and the power of the input signal), and obtain the noise value that exceeds the amplitude threshold. The noise value of the position that does not exceed the amplitude threshold is set to 0.

[0065] Obtain noise values.

[0066]

[0067]

[0068] And record the corresponding index idx, A m This is the peak elimination threshold;

[0069] If x no or x 0p5no If the value is greater than 0, compare x. no and x 0p5no size:

[0070]

[0071]

[0072]

[0073] The above formula calculates x. n_flt (idx), p s (idx) and x ang (idx).

[0074] Step 4: When the number of RBs in the system bandwidth configuration is very small (less than the bandwidth threshold of 10 PRBs), perform a moving average on the noise. Assuming the moving average window length is k, the specific method is as follows:

[0075] For x n_flt (idx)noise is used to perform a moving average filter:

[0076]

[0077] In the formula, x flt Here, x represents the noise value after moving average filtering, k is the moving average window length, and x is the noise value after moving average filtering. n_flt The input noise amplitude.

[0078] Step 5: For x flt Perform a sliding window test:

[0079]

[0080] p pos =p pos ∪idx;

[0081] In the formula, x flt The noise value is the result of moving average filtering, and L is the sliding window detection length.

[0082] Step Six: Calculate the complex value of the peak point based on the amplitude and phase angle:

[0083]

[0084] In the formula, 1i is the imaginary part.

[0085] Step Seven: x n_in p in (idx) s (idx) = 0 passes through the low-pass filter filter0, x n_in p in (idx) s (idx) = 1 passes through the low-pass filter filter1;

[0086] If p s (idx) = 0 and idx ∈ p pos ,but

[0087] x ot (k,idx)=x n_in (k)×C0(j), j=0,...,M-1

[0088] k = idx + j

[0089] In the formula, x ot (.) represents the value obtained by multiplying the complex value of the noise by the value of the filter coefficient, M is the coefficient length of the filter, and C0(.) is the value of the filter0 coefficient;

[0090] If p s (idx) = 1 and idx ∈ p pos ,but

[0091] x ot (k,idx)=x n_in (k)×C1(j), j=0,...,M-1

[0092] k = idx + j

[0093]

[0094] In the formula, C1(.) is the coefficient value of filter1, and x0(.) is the sum of the complex values ​​of noise at different positions multiplied by the filter coefficient values;

[0095] Step 8: Subtract x0(idx) from the input signal x(idx):

[0096] x c (idx) = x(idx) - x0(idx)

[0097] Using the above formula, x can be calculated. c (idx).

[0098] Step 9: Repeat steps 1 through 8 until n times, where n is a positive integer, to obtain the final x. c (idx);

[0099] Step 10, x c (idx) passes through an all-pass filter with a delay of 0.25.

[0100] Step 11: Then repeat steps 1-8 n times;

[0101] The peak reduction function is completed at one sampling rate. By performing a sliding filter on the peak value, when the bandwidth of the input signal is very small, a sliding average filter is applied to the peak value, resulting in a very good peak reduction effect.

[0102] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of any of the methods described above.

[0103] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0104] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A low-complexity peak removal method for NR, characterized in that, include: Step 1, input signal Input a digital filter with a delay of 0.5 to obtain ; Step 2, calculate the magnitude of x. Phase angle of x , amplitude and phase angle ; Step 3: Obtain the noise value. ， ， And record the corresponding index idx, A m This is the peak elimination threshold; like or If greater than 0, then compare. and size: ， ， ， Calculated using the above formula , and ; Step 4: When the system bandwidth configuration RB is less than the set bandwidth threshold, perform a moving average filter on the noise. ， In the formula, The signal is the result of the moving average filtering, where k is the moving average window length. The input noise amplitude is idx, and idx is the index. Step 5: [Regarding...] Perform a sliding window test: ， ， In the formula, P pos It records the peak value. The noise value is the merged value, and L is the sliding window detection length; Step 6: Calculate the complex value of the peak point based on the amplitude and phase angle. ; Step 7, Zhongruo The bandwidth is then compressed using the low-pass filter filter0. like The bandwidth is then compressed using low-pass filter filter1; calculate x0(.) is the sum of the complex noise values ​​at different positions multiplied by the filter coefficient values; Step 8: Subtract x0(idx) from the input signal x(idx): x c (idx)=x(idx)-x0(idx) Using the above formula, x can be calculated. c (idx); Step nine: Repeat steps one through eight until n times, where n is a positive integer, to obtain the final x. c (idx); Step 10, x c (idx) passes through an all-pass filter with a delay of 0.

25. Step 11, then repeat steps 1 through 8 until n times.

2. The low-complexity peak removal method for NR according to claim 1, characterized in that, In step two, the magnitude of x is calculated. Phase angle of x , amplitude and phase angle It can be expressed by the following formula: ， In the formula, This indicates finding the modulus of a complex value; This represents the phase angle function for finding complex numbers.

3. The low-complexity peak removal method for NR according to claim 1, characterized in that, In step six, the complex value of the peak point is calculated based on the amplitude and phase angle. It can be expressed by the following formula: ， In the formula, 1i is the imaginary part.

4. The low-complexity peak removal method for NR according to claim 1, characterized in that, Step seven also includes: like and ,but ， In the formula, x ot (.) represents the value obtained by multiplying the complex value of the noise by the value of the filter coefficient, M is the coefficient length of the filter, and C0(.) is the value of the filter0 coefficient; like and ,but ， ； In the formula, C1(.) represents the coefficient value of filter1.

5. A low-complexity peak removal method for NR according to claim 1, characterized in that, In step three, if or If the amplitude does not exceed the set threshold, then... or The noise value is set to 0.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method according to any one of claims 1 to 5.

Images