Self-adaptive peak cancellation method based on signal bandwidth
By adaptively generating peak-clipping pulse functions, the problem that fixed pulse shapes cannot adapt to signal bandwidth in OFDM systems is solved, achieving low sidelobe clipping and reduced computational complexity, thereby improving signal processing efficiency and performance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- GUANGZHOU RUNXIN INFORMATION TECH CO LTD
- Filing Date
- 2025-12-30
- Publication Date
- 2026-04-28
AI Technical Summary
In existing OFDM systems, fixed pulse shapes cannot adapt to different signal bandwidths and dynamically changing signal characteristics, resulting in larger time-domain sidelobes and increased computational complexity. Furthermore, complex signal analysis modules increase system implementation complexity and power consumption.
By obtaining the total bandwidth of the OFDM signal, an adaptive peak-clipping pulse function is generated. The peak-clipping pulse is generated by multiplying the band-limited pulse with the parameterized window function. Only the total bandwidth needs to be known without resolving the subcarrier position, which reduces computational complexity and optimizes the pulse shape.
Generate efficient, low-sidelobe clipping pulses, reduce the number of clipping iterations, improve the performance trade-off between adjacent band power ratio and error vector amplitude, and reduce system latency and complexity.
Smart Images
Figure CN121940256A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wireless communication technology, and in particular to an adaptive peak cancellation method based on signal bandwidth. Background Technology
[0002] In orthogonal frequency division multiplexing (OFDM) systems, peak-to-average power ratio (PAPR) is a key challenge. Peak cancellation (PC) is a widely used CFR technique. Its core idea is to generate a band-limited clipping pulse that matches the amplitude and phase of the original signal at the peak position, and subtract this pulse from the original signal to bring the signal below a set threshold.
[0003] Traditional CFR pulse generation methods are typically based on a fixed, predefined pulse shape, whose time-domain waveform often exhibits significant sidelobe characteristics, such as a Sinc-like shape. However, this method has inherent drawbacks: the fixed pulse shape cannot adapt to different signal bandwidth configurations and dynamically changing signal characteristics. When canceling a dominant peak, a fixed pulse (such as a Sinc function) generates large time-domain sidelobes. These sidelobes may introduce new secondary peaks at other locations in the signal, requiring multiple stages of iterative peak clipping, increasing computational complexity and system latency.
[0004] In existing technologies, some solutions attempt to dynamically generate pulses based on detailed signal characteristics (such as precise subcarrier distribution and active resource block locations) to optimize performance. For example, Xilinx's PC-CFR solution requires knowledge of the subcarrier distribution and obtains the peak clipping filter by replicating the prototype filter. However, this requires a complex signal analysis module, increasing the complexity and power consumption of the system implementation. Summary of the Invention
[0005] In view of the above problems, the present invention is proposed to provide an adaptive peak cancellation method based on signal bandwidth that overcomes or at least partially solves the above problems.
[0006] This invention discloses an adaptive peak cancellation method based on signal bandwidth, the method comprising: Obtain the total bandwidth of the input OFDM signal; Based on the total bandwidth, an adaptive peak-shaving pulse function is generated; Based on the peak clipping pulse function, the input OFDM signal is subjected to peak cancellation processing.
[0007] This invention has the following advantages: The adaptive peak cancellation method based on signal bandwidth of the present invention obtains the total bandwidth of the input OFDM signal, adaptively generates a peak clipping pulse function based on the total bandwidth, and performs peak cancellation processing on the input OFDM signal based on the generated peak clipping pulse function. This invention only requires knowledge of the total bandwidth of the OFDM signal, without needing to analyze detailed information such as the internal subcarrier positions, to generate efficient, low-sidelobe peak clipping pulses. This reduces the number of peak clipping iterations, lowers computational complexity, and improves the performance trade-off between adjacent band power ratio (ACPR) and error vector magnitude (EVM). Attached Figure Description
[0008] Figure 1 This is a flowchart of the steps of an adaptive peak cancellation method based on signal bandwidth provided by the present invention; Figure 2 This is the PC-CFR flowchart provided by the present invention; Figure 3 This is a flowchart of the hardware implementation of the peak clipping filter provided by the present invention; Figure 4 This is a pseudocode diagram illustrating the hardware implementation of the peak clipping filter provided by this invention; Figure 5 This is a time-domain diagram of the signal before and after one iteration of PC-CFR processing according to the present invention. Figure 6 These are time-domain diagrams of the Xilinx scheme before and after PC-CFR processing following one iteration. Figure 7 These are time-domain diagrams of the signal before and after PC-CFR processing using the Xilinx scheme through three iterations; Figure 8 This is a comparison of the spectrum before and after peak reduction in the present invention. Detailed Implementation
[0009] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
[0010] Reference Figure 1 The diagram illustrates a flowchart of an adaptive peak cancellation method based on signal bandwidth provided in an embodiment of the present invention, which may specifically include the following steps: Obtain the total bandwidth of the input OFDM signal; Based on the total bandwidth, an adaptive peak-shaving pulse function is generated; Based on the peak clipping pulse function, the input OFDM signal is subjected to peak cancellation processing.
[0011] In an optional embodiment of the present invention, the total bandwidth of the OFDM signal is obtained directly from the upper-layer communication protocol stack or signal configuration information, or estimated by analyzing the spectral characteristics of the OFDM signal.
[0012] In an optional embodiment of the present invention, the peak-clipping pulse function is adaptively generated based on the total bandwidth, including: An ideal band-limited pulse and a parameterized window function are constructed; the shape control parameters of the window function are adaptively adjusted according to the total bandwidth of the OFDM signal. The peak-shaving pulse function is obtained by multiplying the ideal band-limited pulse by the parameterized window function.
[0013] In an alternative embodiment of the present invention, the ideal band-limited pulse is a Sinc function.
[0014] In an optional embodiment of the present invention, the shape control parameters are determined based on the total bandwidth of the OFDM signal through a preset nonlinear mapping relationship.
[0015] In an optional embodiment of the present invention, the nonlinear mapping relationship is configured as follows: for a larger total bandwidth, a smaller shape control parameter is mapped to generate a clipped pulse with a narrower main lobe in the time domain; for a smaller total bandwidth, a larger shape control parameter is mapped to generate a clipped pulse with stronger sidelobe suppression.
[0016] In an optional embodiment of the present invention, the nonlinear mapping relationship is an inverse proportional relationship.
[0017] In an optional embodiment of the present invention, the window function is a Kaiser window, and the shape control parameter is the shape parameter of the Kaiser window.
[0018] In an optional embodiment of the present invention, after the step of adaptively generating the peak clipping pulse function based on the total bandwidth, the method further includes: The target pulse sequence is obtained based on the generated peak-shaving pulse function; The target pulse sequence is divided into multiple segments; For each sampling point within a segment, a polynomial of a preset order is used for fitting to obtain the polynomial coefficients corresponding to each segment. The polynomial coefficients are stored as a lookup table.
[0019] In an optional embodiment of the present invention, peak cancellation processing is performed on the input OFDM signal based on the peak clipping pulse function, including: The corresponding polynomial coefficients are found in the lookup table based on the input time index, and the peak-shaving pulse is obtained through polynomial calculation. Detect the position and amplitude of peak values exceeding a preset threshold in the input OFDM signal; Based on the peak position and peak amplitude, the peak clipping pulse is shifted and scaled accordingly to obtain the cancellation pulse; The cancel pulse is subtracted from the input OFDM signal to obtain the peak-clipping signal.
[0020] The PC-CFR implementation process is as follows: Figure 2 The process can be roughly divided into three steps: 1. Peak detection; 2. Peak scaling; 3. Peak clipping filter allocation.
[0021] This invention focuses on the design of a peak clipping filter. In the entire CFR process, the peak clipping pulse generator is the core of the algorithm. The peak clipping pulse needs to be designed to have the same frequency band as the input signal; otherwise, it will cause severe out-of-band leakage. The following is the specific process of constructing the peak clipping filter and performing peak cancellation according to this invention: S1: Signal parameter acquisition.
[0022] Obtain the total bandwidth B of the input OFDM signal and the target sampling frequency Fs of the system. The total bandwidth B is directly given by the upper-layer communication protocol stack or signal configuration information, or automatically estimated by analyzing the spectral characteristics of the signal. The target sampling rate Fs must satisfy the Nyquist sampling theorem, i.e., Fs > B.
[0023] S2: Adaptive impulse function generation.
[0024] Based on the total bandwidth B, a peak-shaving pulse function with optimized time-frequency characteristics is dynamically calculated or adaptively selected. This step is the core of the invention, and its specific implementation is as follows: S2.1: Define the basis structure of the impulse function.
[0025] The peak-shaving pulse function From an ideal band-limited pulse With a parameterized window function Generate its mathematical expression by performing dot product:
[0026] in: An ideal band-limited pulse has a flat frequency response within bandwidth B and zero outside the bandwidth. Its preferred time-domain form is a Sinc function, i.e.
[0027] For a parameterized window function, It is its key shape control parameter, used to regulate the pulse. The trade-off between temporal sidelobe attenuation and main lobe width.
[0028] S2.2: Establish bandwidth B and control parameters The mapping relationship.
[0029] The control parameters A pre-defined, non-linear mapping relationship is established with the signal bandwidth B. The design principle of this mapping relationship is: For signals with a large bandwidth B: automatically select or calculate a smaller one. Value. This parameter generates pulses with a narrow main lobe width in the time domain. Narrow pulses can more accurately align with and cancel out sharp peaks in the time domain, avoiding excessive diffusion that could affect surrounding sampling points. They are suitable for scenarios that require high peak positioning accuracy and allow for a certain degree of sidelobes.
[0030] For signals with a smaller bandwidth B: automatically select or calculate a larger one. Value. This parameter generates pulses with excellent sidelobe suppression (large sidelobe attenuation). Wide pulses minimize the risk of introducing secondary peaks elsewhere when canceling a primary peak, thus reducing the number of iterations required for CFR processing. This makes them suitable for scenarios with extremely stringent requirements for out-of-band leakage and signal distortion.
[0031] A preferred mapping relationship is an inverse proportional relationship. ,in Let be a constant determined through simulation or experiment. More complex mappings can be achieved using lookup tables or piecewise functions.
[0032] A specific example is as follows: Design goal: To design CFR pulses for a 5G NR base station transmitter that supports channel bandwidths of 20MHz, 40MHz, 60MHz, 80MHz, and 100MHz.
[0033] Assumption: Sampling rate Fs: 122.88 MHz (This is a commonly used sampling rate for 5G base stations, approximately 1.2288 times that of 100MHz bandwidth) Pulse length: 13 sampling points (symmetrical, with the peak value at the center and 6 points on each side). Core pulse: (Ideal low-pass pulse) Window function: Kaiser window (due to its shape parameter) It can precisely control the trade-off between main lobe width and side lobe attenuation (industry standard). Step 1: Determine the range of bandwidth B The signal bandwidth B varies, but its maximum value will not exceed the Nyquist frequency. Therefore, the range of B is roughly between 10MHz and 60MHz.
[0034] Step 2: Determine the Kaiser window parameters Scope Based on Kaiser window design experience: (Rectangular window): The main lobe is the narrowest (~2.0 DFT bins), but the side lobe attenuation is poor (~-13 dB).
[0035] The main lobe is relatively wide (~2.8 bins), and the side lobes have good attenuation (~-40 dB).
[0036] The main lobe is very wide (~3.6 bins), and the side lobes have excellent attenuation (~-60 dB).
[0037] For CFR, a larger dynamic range is required. Therefore, The range is set between 3 and 8.
[0038] Step 3: Establish mapping relationship
[0039] Large bandwidth B -> Small -> Narrow main lobe, low sidelobe suppression: Because the bandwidth of a large signal is relatively wide, the requirements for the transition band of the filter are relatively relaxed, and a slightly higher sidelobe can be tolerated in exchange for time domain accuracy.
[0040] Small bandwidth B -> Large -> Wide main lobe, high sidelobe suppression: Small bandwidth signals must have extremely steep filtering edges to prevent energy leakage to adjacent channels. Therefore, extremely high sidelobe suppression is required, even at the cost of some time-domain main lobe width.
[0041] Using an inverse proportional relationship:
[0042] Step 4: Determine constants through simulation
[0043] Need to find one This makes it possible for B to be within its range. It can reasonably fall within the interval [3, 8].
[0044] When B = 60MHz (maximum bandwidth), we hope
[0045] 3 = k / 60 => k = 180 When B = 20MHz (smaller bandwidth), we hope
[0046] 8 = k / 20 => k = 160 The value is between 160 and 180. A middle value can be chosen. And fine-tuning was performed. The final mapping relationship is as follows:
[0047] The unit of B is MHz.
[0048] S2.3: Specific implementation of the window function.
[0049] The window function Various functions that can achieve the above-mentioned parameterized control can be selected. The Kaiser window used in this invention is: Kaiser window: Its expression is ,in It is a zero-order modified Bessel function. The control parameters... It can be directly mapped to the shape parameters of the Kaiser window. . The larger the value, the greater the side lobe attenuation and the wider the main lobe.
[0050] S3: Hardware implementation of the peak clipping filter.
[0051] Although the peak clipping filter designed in this way only requires bandwidth information to be successfully designed, the complexity of direct implementation is too high because it involves complex hardware calculations such as Bessel functions, while directly storing the filter coefficients generated by MATLAB simulation consumes too much storage resources.
[0052] like Figure 3 The hardware implementation flowchart of the peak clipping filter is shown below. This invention employs a piecewise polynomial approximation method. First, the ideal peak clipping filter is calculated in software using an accurate mathematical expression. Then, the pulse time-domain sequence is divided into several segments, each containing several sampling points. For the sampling points of each segment, a polynomial fitting is used to approximate the target pulse. The polynomial coefficients of each segment are obtained through least squares fitting, and these coefficients form a lookup table for hardware implementation. Simultaneously, the performance of the polynomial approximation can be evaluated in software to ensure that performance is not degraded while optimizing storage resources. Taking a second-order polynomial as an example, the pseudocode diagram of the hardware implementation of the peak clipping filter is shown below. Figure 4 Its solution function is as follows:
[0053] The coefficients a, b, and c can be obtained using the built-in MATLAB function polyfit(x,y,n). For the desired target pulse segment, The normalized time is divided according to the pulse length, and n is the polynomial order.
[0054] The hardware first normalizes the input time index to facilitate polynomial calculation; based on the current time index, it determines the segment to which it belongs and reads the polynomial coefficients of the corresponding segment through a lookup table; the polynomial calculation module is implemented, using fixed-point multipliers and adders, and a pipelined design to improve the clock frequency; and an approximate pulse value is output.
[0055] This method is mainly applied when the target pulse filter is long, and the number of segments is denoted as... The order of a polynomial is denoted as The filter length is denoted as The conditional function is as follows:
[0056] This method only has practical engineering significance when it holds true. According to the results obtained from MATLAB simulation, when the length of the peak clipping filter is 30, the second-order polynomial coefficients can be solved in 6 segments. At this time, the pulse obtained from solving the coefficients is similar to the target pulse, while significantly reducing the need for adders and multipliers, thus saving space resources.
[0057] S4: Peak cancellation processing.
[0058] The digital peak clipping pulse sequence For routine peak cancellation CFR processing: Detecting the input signal For peak values exceeding a preset threshold, the pulse sequence will be scaled (the scaling factor is determined by the peak amplitude) and time-shifted (to align the peak positions). Subtracting from the original signal yields the peak-shaving signal. ,in Scaling factor This is the peak position. Figure 5 This is a time-domain diagram of the signal before and after one iteration of PC-CFR processing according to the present invention. Figure 6 These are time-domain diagrams of the Xilinx scheme before and after PC-CFR processing following one iteration. Figure 7 These are time-domain diagrams of the signal before and after PC-CFR processing using the Xilinx scheme through three iterations; Figure 8 This is a comparison of the spectrum before and after peak reduction in the present invention.
[0059] The present invention has the following beneficial effects: (1) Reduced complexity: There is no need to parse OFDM symbol structure or subcarrier mapping information. It only relies on the easily obtainable system parameter of bandwidth, which greatly simplifies system design.
[0060] (2) Improve performance: Through bandwidth-adaptive pulse shape optimization, the generated pulse has smaller time-domain side lobes, which can effectively reduce the number of iterations required for CFR processing, thereby reducing computational latency.
[0061] (3) Improve performance trade-offs: By optimizing pulse energy distribution, better in-band signal quality (EVM) can be obtained under the same adjacent band power ratio (ACPR) performance, or better ACPR performance can be achieved under the same EVM requirements.
[0062] It should be noted that, for the sake of simplicity, the method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments of the present invention are not limited to the described order of actions, because according to the embodiments of the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to the embodiments of the present invention.
[0063] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0064] The various embodiments in this specification are described in a related manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0065] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.
Claims
1. An adaptive peak cancellation method based on signal bandwidth, characterized in that, The method includes: Obtain the total bandwidth of the input OFDM signal; Based on the total bandwidth, an adaptive peak-shaving pulse function is generated; Based on the peak clipping pulse function, the input OFDM signal is subjected to peak cancellation processing.
2. The method according to claim 1, characterized in that, The total bandwidth of the OFDM signal is obtained directly from the upper-layer communication protocol stack or signal configuration information, or estimated by analyzing the spectral characteristics of the OFDM signal.
3. The method according to claim 1, characterized in that, Based on the total bandwidth, an adaptive peak-shaving pulse function is generated, including: An ideal band-limited pulse and a parameterized window function are constructed; the shape control parameters of the window function are adaptively adjusted according to the total bandwidth of the OFDM signal. The peak-shaving pulse function is obtained by multiplying the ideal band-limited pulse by the parameterized window function.
4. The method according to claim 3, characterized in that, The ideal band-limited pulse is the Sinc function.
5. The method according to claim 3, characterized in that, The shape control parameters are determined based on the total bandwidth of the OFDM signal through a preset nonlinear mapping relationship.
6. The method according to claim 5, characterized in that, The nonlinear mapping relationship is configured such that, for a larger total bandwidth, a smaller shape control parameter is obtained to generate a clipped pulse with a narrower main lobe in the time domain. For a smaller total bandwidth, a larger shape control parameter is obtained by mapping to generate a clipped pulse with stronger sidelobe suppression.
7. The method according to claim 5, characterized in that, The nonlinear mapping relationship is an inverse proportional relationship.
8. The method according to claim 3, characterized in that, The window function is a Kaiser window, and the shape control parameters are the shape parameters of the Kaiser window.
9. The method according to claim 1, characterized in that, Following the step of adaptively generating the peak-clipping pulse function based on the total bandwidth, the method further includes: The target pulse sequence is obtained based on the generated peak-shaving pulse function; The target pulse sequence is divided into multiple segments; For each sampling point within a segment, a polynomial of a preset order is used for fitting to obtain the polynomial coefficients corresponding to each segment. The polynomial coefficients are stored as a lookup table.
10. The method according to claim 9, characterized in that, Based on the peak-clipping pulse function, peak cancellation processing is performed on the input OFDM signal, including: The corresponding polynomial coefficients are found in the lookup table based on the input time index, and the peak-shaving pulse is obtained through polynomial calculation. Detect the position and amplitude of peak values exceeding a preset threshold in the input OFDM signal; Based on the peak position and peak amplitude, the peak clipping pulse is shifted and scaled accordingly to obtain the cancellation pulse; The cancel pulse is subtracted from the input OFDM signal to obtain the peak-clipping signal.