Channel overlapping signal fusion method and system based on cross-channel background comparison
The channel overlap signal fusion method based on cross-channel background comparison solves the problem of repeated signal detection in multi-filter channelized receivers, thereby improving the accuracy of signal detection and processing efficiency. It is applicable to hardware platforms such as FPGA.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NANJING RES INST OF ELECTRONICS TECH
- Filing Date
- 2026-02-11
- Publication Date
- 2026-05-15
AI Technical Summary
The problem of repeated signal detection caused by channel frequency overlap in multi-channel filtered receivers increases the complexity of signal sorting and computational resource consumption.
A channel overlap signal fusion method based on cross-channel background comparison is adopted. By calculating the threshold using the background window of adjacent channels during the channelization stage, the redundant signals generated by channel overlap are directly fused to distinguish between real signals and redundant signals.
Signal fusion is performed directly during the signal detection stage, which improves the accuracy and efficiency of signal detection, reduces computational complexity, and is suitable for implementation on hardware platforms such as FPGA.
Smart Images
Figure CN122053016A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital signal processing technology, and more specifically, to a method and system for channel overlap signal fusion based on cross-channel background comparison. Background Technology
[0002] Multi-frequency filtering channelized receivers are widely used due to their ability to process wideband signals in parallel. This technology divides a wideband into multiple sub-channels using a set of overlapping frequency-domain filters. However, the design of this filter introduces an inherent technical problem: a frequency signal located at the edge of two channels will be received by both channels simultaneously, resulting in two similar signal responses at the output in the same distance cell of two adjacent channels. The same signal is repeatedly detected in multiple channels, increasing the complexity of signal sorting. Subsequent processing systems need to perform additional correlation and deduplication processing on these repeated signals, consuming computational resources. Summary of the Invention
[0003] The purpose of this invention is to overcome the problem of repeated signal detection caused by channel frequency overlap in multi-channel filtered channelized receivers, and to provide a channel overlap signal fusion method and system based on cross-channel background comparison, which can directly fuse redundant signals generated by channel overlap during the detection stage, thereby improving the accuracy of signal detection and processing efficiency.
[0004] To achieve the above objectives, the technical solution adopted by the present invention is as follows: a channel overlap signal fusion method based on cross-channel background comparison, comprising the following steps:
[0005] S1: Channelization, using multi-filter channelization technology to evenly divide the wideband signal into N parallel narrowband channels according to a preset frequency interval;
[0006] S2: Construct a background window. For the current distance cell in the current channel, take the current distance cells of its two adjacent channels, as well as its previous NLen point as the front window and the next N points as the back window. NLen≥1. Use the distance cells of these adjacent channels as the background window.
[0007] S3: Threshold calculation: Calculate the maximum value of the amplitude of all distance points within the background window, and use this maximum value as the threshold to determine whether it needs to be merged.
[0008] S4: Fusion decision: Compare the current distance gate amplitude of the current channel with the fusion threshold. If it is smaller than the threshold, it is determined that the point is a signal leakage from another channel and needs to be fused. If it is larger than the threshold, it is determined that the point is a normal signal point and does not need to be fused.
[0009] S5: Fusion processing. For points that need to be fused, their amplitude is replaced with noise or 0. For points that do not need to be fused, no processing is performed.
[0010] In a preferred embodiment of the present invention, in S1:
[0011] Each channel has M sampling points, let This represents the amplitude or power value of the k-th channel at the i-th distance cell, where... Indicates the channel index. This represents the distance gate index.
[0012] In a preferred embodiment of the present invention, in step S2:
[0013] The current distance cell in the current channel is denoted as... Background window is recorded as Represented by sets:
[0014] .
[0015] In a preferred embodiment of the present invention, in step S3:
[0016] Find the maximum amplitude of all distance points within the background window. :
[0017]
[0018] That is:
[0019]
[0020] Maximum value m times the threshold used to determine whether fusion is necessary is denoted as . :
[0021] , m=1.
[0022] In a preferred embodiment of the present invention, in step S4:
[0023] Compare the current channel's current range gate amplitude with the fusion threshold:
[0024]
[0025] If the gate amplitude is smaller than the fusion threshold, the point is determined to be a signal leaked from another channel and needs to be fused, and is recorded as 0. If the gate amplitude is larger than the fusion threshold, the point is determined to be a normal signal point, and is recorded as 1, and does not need to be fused.
[0026] In a preferred embodiment of the present invention, in step S5:
[0027] For points that need to be merged, their amplitudes are replaced with noise or 0; for points that do not need to be merged, no processing is performed.
[0028] ;
[0029] If points that should not have been merged are mistakenly identified as needing to be merged after merging, compensation will be provided for the lost points:
[0030] ;
[0031] If, after fusion, there are burrs that need to be fused but are mistakenly judged as not needing to be fused, these burrs will be removed using a narrow pulse.
[0032] The present invention also provides a channel overlap signal fusion system based on cross-channel background comparison, comprising:
[0033] The channelization module uses multiple filtering channelization techniques to evenly divide the wideband signal into N parallel narrowband channels according to a preset frequency interval;
[0034] The background window module is constructed by taking the current distance unit in the current channel as the front window, the two adjacent current distance units of the current channel, and the previous NLen point as the back window, and using the distance units of these adjacent channels as the background window.
[0035] The threshold calculation module calculates the maximum value of the amplitude of all distance points within the background window. The m times of this maximum value is used as the threshold to determine whether it needs to be fused. The default m=1.
[0036] The fusion decision module compares the current distance threshold of the current channel with the fusion threshold. If it is smaller than the fusion threshold, it determines that the point is a signal leaked from another channel and needs to be fused. If it is larger than the fusion threshold, it determines that the point is a normal signal point and does not need to be fused.
[0037] The fusion processing module replaces the amplitude of points that need to be fused with noise or 0, and does not process points that do not need to be fused.
[0038] As described above, the technical solution adopted in this invention has the following beneficial effects:
[0039] 1. This invention can directly solve the fusion problem in the detection stage: no complicated post-processing is required, and the fusion of overlapping signals is completed at the same time as signal detection, resulting in high processing efficiency;
[0040] 2. The present invention has accurate fusion judgment. By using the background of adjacent channels as a reference, it can accurately distinguish between redundant signals caused by channel overlap and real effective signals.
[0041] 2. The algorithm structure of this invention is clear, the computational complexity is low, and it is easy to implement on hardware platforms such as FPGA using hardware languages such as Verilog and VHDL. Attached Figure Description
[0042] Figure 1 This is a schematic diagram of the channel overlap signal fusion method based on cross-channel background comparison of the present invention;
[0043] Figure 2 This is a schematic diagram of the subband characteristics of the filter of the present invention;
[0044] Figure 3 This is a schematic diagram of the cross-channel background window of the present invention (NLen=2);
[0045] Figure 4 This is a schematic diagram illustrating the simulation effect of the present invention. Detailed Implementation
[0046] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0047] The embodiments provided by the present invention will be described in detail below:
[0048] like Figure 1-3 As shown, this embodiment provides an amplitude estimate based on real-time pipelined signaling between sub-channels. For each range cell in each channel, reference cells surrounding the corresponding positions in its two adjacent channels are used as background windows to calculate the amplitude estimate of this window. This estimate is then used as a threshold benchmark to determine whether the signal of the current range cell in the current channel should be fused. The main steps include:
[0049] S1: Channelization, using multi-filter channelization technology to evenly divide the wideband signal into N parallel narrowband channels according to a preset frequency interval;
[0050] S2: Construct a background window. For the current distance cell in the current channel, take the current distance cells of its two adjacent channels, as well as its previous NLen point as the front window and the next N points as the back window. NLen≥1. Use the distance cells of these adjacent channels as the background window.
[0051] S3: Threshold calculation: Calculate the maximum value of the amplitude of all distance points within the background window, and use this maximum value as the threshold to determine whether it needs to be merged.
[0052] S4: Fusion decision: Compare the current distance gate amplitude of the current channel with the fusion threshold. If it is smaller than the threshold, it is determined that the point is a signal leakage from another channel and needs to be fused. If it is larger than the threshold, it is determined that the point is a normal signal point and does not need to be fused.
[0053] S5: Fusion processing. For points that need to be fused, their amplitude is replaced with noise or 0. For points that do not need to be fused, no processing is performed.
[0054] As a preferred embodiment, in step S1:
[0055] Each channel has M sampling points, let This represents the amplitude or power value of the k-th channel at the i-th distance cell, where... Indicates the channel index. This represents the distance gate index.
[0056] As a preferred embodiment, in step S2:
[0057] The current distance cell in the current channel is denoted as... Background window is recorded as Represented by sets:
[0058] .
[0059] As a preferred embodiment, in step S3:
[0060] Find the maximum amplitude of all distance points within the background window. :
[0061]
[0062] That is:
[0063]
[0064] Maximum value m times the threshold used to determine whether fusion is necessary is denoted as . :
[0065] , m=1.
[0066] As a preferred embodiment, in step S4:
[0067] Compare the current channel's current range gate amplitude with the fusion threshold:
[0068]
[0069] If the gate amplitude is smaller than the fusion threshold, the point is determined to be a signal leaked from another channel and needs to be fused, and is recorded as 0. If the gate amplitude is larger than the fusion threshold, the point is determined to be a normal signal point, and is recorded as 1, and does not need to be fused.
[0070] As a preferred embodiment, in step S5:
[0071] For points that need to be merged, their amplitudes are replaced with noise or 0; for points that do not need to be merged, no processing is performed.
[0072] ;
[0073] If points that should not have been merged are mistakenly identified as needing to be merged after merging, compensation will be provided for the lost points:
[0074] ;
[0075] If, after fusion, there are burrs that need to be fused but are mistakenly judged as not needing to be fused, these burrs will be removed using a narrow pulse.
[0076] like Figure 4 As shown, this embodiment simulates a 480M broadband sampling IQ (real and imaginary parts). The sampling includes a point frequency signal with a pulse width of 10µs, a period of 50µs, and a frequency of 12M. After channelization, this signal falls into channels 9 and 10. The real signal is in channel 9, while channel 10 is a redundant signal due to the cross-banding of the filter.
[0077] This analog input is fed into the channelization module, and the degree-term filtering channelization technology is used to evenly divide the wideband signal into N=16 parallel narrowband channels according to the preset frequency interval. Each channel has a sampling rate of 60M and a bandwidth of 30M.
[0078] The background window module is constructed. For the current distance cell in the current channel, the current distance cells of its two adjacent channels, as well as the previous NLen=2 point and the next NLen point are used as the front window and the back window, and the distance cells of these adjacent channels are used as the background window.
[0079] The threshold calculation module calculates the maximum value of the amplitude of all distance points within the background window. The m=1 times of this maximum value is used as the threshold to determine whether it needs to be fused.
[0080] The fusion decision module compares the current distance threshold of the current channel with the fusion threshold. If it is smaller than the threshold, the point is determined to be a signal leaked from another channel and needs to be fused. If it is larger than the threshold, the point is determined to be a normal signal point and does not need to be fused.
[0081] The fusion processing module replaces the amplitude of points that need to be fused with noise or 0, and does not process points that do not need to be fused.
[0082] Processing simulation effects such as Figure 4 Before fusion, both channel 9 and channel 10 could detect signals. After fusion, only channel 9 could detect signals, while channel 10 could not, thus achieving the purpose of fusion.
[0083] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A channel overlap signal fusion method based on cross-channel background comparison, characterized in that, Includes the following steps: S1: Channelization, using multi-filter channelization technology to evenly divide the wideband signal into N parallel narrowband channels according to a preset frequency interval; S2: Construct a background window. For the current distance cell in the current channel, take the current distance cells of its two adjacent channels, as well as its previous NLen point as the front window and the next N points as the back window. NLen≥1. Use the distance cells of these adjacent channels as the background window. S3: Threshold calculation: Calculate the maximum value of the amplitude of all distance points within the background window, and use this maximum value as the threshold to determine whether it needs to be merged. S4: Fusion decision: Compare the current distance gate amplitude of the current channel with the fusion threshold. If it is smaller than the threshold, it is determined that the point is a signal leakage from another channel and needs to be fused. If it is larger than the threshold, it is determined that the point is a normal signal point and does not need to be fused. S5: Fusion processing. For points that need to be fused, their amplitude is replaced with noise or 0. For points that do not need to be fused, no processing is performed.
2. The channel overlap signal fusion method based on cross-channel background comparison according to claim 1, characterized in that, In S1: Each channel has M sampling points, let This represents the amplitude or power value of the k-th channel at the i-th distance cell, where... Indicates the channel index. This represents the distance gate index.
3. The channel overlap signal fusion method based on cross-channel background comparison according to claim 1, characterized in that, In S2: The current distance cell in the current channel is denoted as... Background window is recorded as Represented by sets: 。 4. The channel overlap signal fusion method based on cross-channel background comparison according to claim 1, characterized in that, In S3: Find the maximum amplitude of all distance points within the background window. : ； That is: ； Maximum value m times the threshold used to determine whether fusion is necessary is denoted as . : ，m=1。 5. The channel overlap signal fusion method based on cross-channel background comparison according to claim 1, characterized in that, In S4: Compare the current channel's current range gate amplitude with the fusion threshold: ； If the gate amplitude is smaller than the fusion threshold, the point is determined to be a signal leaked from another channel and needs to be fused, and is recorded as 0. If the gate amplitude is larger than the fusion threshold, the point is determined to be a normal signal point, and is recorded as 1, and does not need to be fused.
6. The channel overlap signal fusion method based on cross-channel background comparison according to claim 1, characterized in that, In S5: For points that need to be merged, their amplitudes are replaced with noise or 0; for points that do not need to be merged, no processing is performed. ； If points that should not have been merged are mistakenly identified as needing to be merged after merging, compensation will be provided for the lost points: ； If, after fusion, there are burrs that need to be fused but are mistakenly judged as not needing to be fused, these burrs will be removed using a narrow pulse.
7. A channel overlap signal fusion system based on cross-channel background comparison for implementing the method of any one of claims 1-6, characterized in that, include: The channelization module uses multiple filtering channelization techniques to evenly divide the wideband signal into N parallel narrowband channels according to a preset frequency interval; The background window module is constructed by taking the current distance unit in the current channel as the front window, the two adjacent current distance units of the current channel, and the previous NLen point as the back window, and using the distance units of these adjacent channels as the background window. The threshold calculation module calculates the maximum value of the amplitude of all distance points within the background window. The m times of this maximum value is used as the threshold to determine whether it needs to be fused. The default m=1. The fusion decision module compares the current distance threshold of the current channel with the fusion threshold. If it is smaller than the fusion threshold, it determines that the point is a signal leaked from another channel and needs to be fused. If it is larger than the fusion threshold, it determines that the point is a normal signal point and does not need to be fused. The fusion processing module replaces the amplitude of points that need to be fused with noise or 0, and does not process points that do not need to be fused.