Dynamic shift quantization modulation noise suppression method and device for separating strong and weak signals
By employing a dynamic shift quantization modulation method with tiered perturbation noise suppression, the problem of strong direct waves masking weak echoes in non-cooperative signal processing is solved. This method achieves low-complexity, high-fidelity separation of strong and weak signals, improves the signal-to-noise ratio and phase accuracy, and is applicable to radar, communication, and medical imaging.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- THE 20TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORP
- Filing Date
- 2025-09-29
- Publication Date
- 2026-07-14
AI Technical Summary
Existing technologies in non-cooperative signal processing suffer from hardware dependence and insufficient adaptability to complex environments, quantization noise drowning out weak signals, contradictions between dynamic range and real-time performance, and waste of strong signal resources, making it difficult to efficiently separate strong direct waves and weak echoes.
A dynamic shift quantization modulation method with tiered perturbation noise suppression is adopted. Through symbol expansion, dynamic parameter calculation, staged perturbation injection and signal separation, high-fidelity separation of strong and weak signals is achieved, including symbol expansion, sliding window statistics, perturbation injection and signal recovery.
It achieves low-complexity, high-fidelity, and real-time strong and weak signal separation, improves the signal-to-noise ratio by ≥40dB, optimizes the phase error by ≤0.8°, adapts to 86dB instantaneous fluctuations, and is suitable for scenarios such as radar detection, communication anti-interference, and medical imaging.
Smart Images
Figure CN121256294B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of signal processing technology, to non-cooperative signal processing under external radiation source detection systems, and particularly to a method and apparatus for separating strong and weak signals by dynamic shift quantization modulation with graded perturbation noise suppression. Background Technology
[0002] In electromagnetic sensing scenarios, received signals typically consist of strong direct waves (directly from the transmitter) and weak echoes (reflected by the target or scattered by the environment). The dynamic range difference between the two exceeds 60 dB (e.g., the strong signal amplitude is ±32767, while the weak signal is only ±328). Strong direct waves account for over 90% of the signal energy, masking weak echoes and leading to the loss of target information. Weak echoes contain target characteristics (such as range, velocity, and angle), but their amplitude is extremely low, making them easily obscured by noise and quantization errors. For example, in external radiation source radars using broadcast / communication signals as illumination sources, the dynamic range difference between strong direct waves and weak target echoes in the received signal often exceeds 80 dB; in electronic reconnaissance, strong enemy jamming signals are mixed with weak friendly communication signals, making separation difficult using traditional methods.
[0003] Achieving efficient and high-fidelity separation of strong direct wave and weak echo signals is fundamental for subsequent effective demodulation, localization, identification, and tracking. Three common approaches are employed: First, the reference antenna method, which uses an independent reference antenna channel to acquire a "clean" strong direct wave signal and then eliminates the strong signal component through adaptive filtering or cancellation techniques to extract the weak echo. Second, the transform domain separation method, which transforms the signal to the frequency domain, time-frequency domain (such as short-time Fourier transform), or sparse representation domain, utilizing the differences in the distribution of strong and weak signals in different domains for separation. Third, the blind source separation method, based on independent component analysis (ICA) or principal component analysis (PCA), assumes that strong and weak signals are statistically independent and achieves separation through matrix decomposition.
[0004] Comparing the three methods: the reference antenna method has relatively high cancellation accuracy (SNR improvement > 20dB), but it relies on additional hardware, increasing system complexity, cost, and environmental sensitivity with poor adaptability. Multipath interference leads to phase inconsistency, which degrades cancellation performance. The transform domain separation method suppresses strong signal bands in the frequency domain or time-frequency domain (e.g., FFT zeroing). Frequency domain processing is intuitive and suitable for narrowband signals, but computation is complex, real-time performance is poor, and phase consistency is difficult to guarantee. The blind source separation method does not require prior information and has strong theoretical universality, but the assumptions are stringent (strong / weak signals may be related due to multipath effects). In real-world scenarios, its performance is unstable, and iterative convergence is slow with large delays (> 10ms).
[0005] Another method is time-domain adaptive filtering separation, including the LMS / NLMS and RLS algorithms. These methods utilize adaptive filters to suppress strong signals and extract weak signals by adjusting the filter coefficients. The RLS algorithm, relatively speaking, converges faster and is suitable for real-time processing. For stable mixed signals, the weak signal separation effect is acceptable, but for changing signals, the effect is poor, and the computational resources are very high.
[0006] Disadvantages and shortcomings of existing technologies
[0007] Hardware dependency and insufficient adaptability to complex environments: Existing technologies rely on independent reference antennas, floating-point operations, or complex frequency domain transformations, resulting in high hardware costs, poor real-time performance, and performance degradation in scenarios with multipath interference and signal abrupt changes. For example, the reference antenna method requires additional hardware (increasing costs by more than 30%), and multipath interference causes phase mismatch; the reference antenna method requires the additional deployment of high-isolation antennas, and in complex terrain (such as urban multipath environments), when the antenna spacing error exceeds λ / 10 (one-tenth of the wavelength), the cancellation performance degrades by more than 10dB; the frequency domain separation method (FFT / STFT) is computationally complex (delay > 1ms), and cannot meet the real-time requirements of radar pulse level.
[0008] The quantization noise's effect on weak signals: When the amplitude of a weak signal is close to the quantization noise level (e.g., a quantization step size of 1 in a 16-bit ADC), traditional methods amplify the noise simultaneously when amplifying the weak signal, resulting in insufficient improvement in the signal-to-noise ratio (SNR) after separation. For example, RLS filtering can only suppress noise by about 15dB, and the weak signal is still submerged.
[0009] The conflict between dynamic range and real-time performance: Signal dynamic range fluctuates drastically (e.g., radar echoes change instantaneously by 40-70 dB), and traditional adaptive algorithms (such as LMS / RLS) require iterative convergence (delay > 200 μs), making it impossible to balance real-time performance and accuracy. For example, fixed-parameter designs cannot adapt to sudden changes in dynamic range (e.g., a sudden appearance of a target causing a jump in signal amplitude); real-time parameter adjustment algorithms (such as sliding window statistics) have high computational complexity and are difficult to deploy in low-power hardware.
[0010] The problem of wasting resources on strong signals: Existing methods focus on "suppressing strong signals," which leads to the loss of phase and amplitude information of the separated strong signals, making them unusable for subsequent tasks such as synchronization demodulation and reference signal generation. For example, strong signals cannot be reused after being canceled or zeroed in the frequency domain; phase truncation errors (such as setting low-level bits to zero in a right shift operation) cause distortion of strong signals. Summary of the Invention
[0011] This application provides a method and apparatus for separating strong and weak signals using dynamic shift quantization modulation with graded perturbation noise suppression. It provides a low-complexity, high-fidelity, and real-time strong-weak mixed signal separation technology, solving the problem of high-precision separation of extreme signal-to-noise ratio mixed signals in non-cooperative signal processing, which faces the prominent challenge of "strong direct waves masking weak echoes". It provides an efficient and universal mixed signal separation method for fields such as radar detection, communication anti-interference, and electronic reconnaissance.
[0012] This application provides a method for separating non-cooperative strong and weak signals in dynamic shift quantization modulation, including:
[0013] Sign extension of the input signal;
[0014] For the symbol-extended signal, the maximum value Amax and minimum value Amin of the signal are statistically analyzed using a specified sliding window, and the required parameters are dynamically calculated.
[0015] For the symbol-spread signal, a first perturbation d1 is injected, and according to the calculated parameters, the symbol-spread signal injected with the first perturbation d1 is shifted left by m bits to amplify it, and a second perturbation d2 is injected into the amplified signal.
[0016] For the signal injected with the second disturbance d2, the strong signal component is extracted by shifting it right by n bits and left by n bits in sequence, and the weak signal is separated by subtraction.
[0017] The magnitude is recovered by shifting strong and weak signal components to the right by m bits.
[0018] Strong signal components are symmetrically saturated and truncated, while weak signals are asymmetrically truncated and output.
[0019] This application provides a non-cooperative strong / weak signal separation device for dynamic shift quantization modulation, comprising:
[0020] The sign extension module is configured to extend the sign of the input signal;
[0021] The dynamic parameter calculation module uses a specified sliding window to statistically analyze the maximum value Amax and minimum value Amin of the input signal after sign expansion, and dynamically calculates the required parameters.
[0022] The shift amplification module amplifies the sign-extended signal by shifting it left by m bits based on the calculated parameters.
[0023] The staged perturbation injection module is used to inject a first perturbation d1 and a second perturbation d2 into the signal before and after the sign-extended signal is amplified by shifting it left by m bits.
[0024] The strong signal extraction module extracts strong signal components by shifting the signal right by n bits and left by n bits.
[0025] The weak signal separation module separates weak signals through subtraction.
[0026] The magnitude recovery module recovers the magnitude by shifting the strong and weak signal components to the right by m bits.
[0027] The output module uses symmetrical saturation truncation for strong signal components and asymmetrical truncation for weak signals.
[0028] This application provides a low-complexity, high-fidelity, and real-time strong-weak mixed signal separation technology, which solves the problem of high-precision separation of extreme signal-to-noise ratio mixed signals in non-cooperative signal processing, which faces the prominent challenge of "strong direct waves masking weak echoes". It provides an efficient and universal mixed signal separation method for fields such as radar detection, communication anti-jamming, and electronic reconnaissance.
[0029] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description
[0030] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of this application. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:
[0031] Figure 1 This is another flowchart illustrating the non-cooperative strong / weak signal separation method according to an embodiment of this application;
[0032] Figure 2 This is a schematic diagram of the staged disturbance injection DSQM process for the non-cooperative strong and weak signal separation device in this application embodiment;
[0033] Figure 3 This is a schematic diagram of the disturbance dynamic start / stop control state machine transition of the non-cooperative strong / weak signal separation device in this embodiment of the application. Detailed Implementation
[0034] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
[0035] This application proposes a method for separating strong and weak signals in dynamic shift quantization modulation with graded perturbation noise suppression, such as... Figure 1 As shown, it includes the following steps:
[0036] In step S101, the input signal is sign-extended. In some embodiments, sign-extending the input signal includes reducing the bit width from B... in Extend to B ext =2×B in The high-order bit is filled with the sign bit.
[0037] In step S102, for the symbol-extended signal, the maximum value Amax and minimum value Amin of the signal are statistically analyzed using a specified sliding window, and the required parameters are dynamically calculated;
[0038] In step S103, a first perturbation d1 is injected after the symbol expansion, with an amplitude |d1|=clamp(2^(mn),1,Amin / 8) and a bit width of low log2(Amin)+3 bits.
[0039] In step S104, the extended signal injected with the first disturbance d1 is shifted left by m bits and amplified according to the calculated parameters;
[0040] In step S105, a second perturbation d2 is injected after amplification, with an amplitude of |d2|=clamp(2^(mn),1,Amin·2^m / 8) and a bit width of low m+log2(Amin)+3 bits.
[0041] In step S106, the strong signal components are extracted by shifting right by n bits and left by n bits in sequence.
[0042] In step S107, the weak signal is separated by subtraction, and the strong signal component and the weak signal component are right-shifted by m bits to restore the magnitude.
[0043] In step S108, the strong signal component is symmetrically saturated and truncated, while the weak signal is output with asymmetric truncation.
[0044] This application solves the core contradictions in traditional methods, such as "strong signal suppression leading to weak signal phase distortion", "conflict between quantization noise and dynamic range", and "incompatibility between hardware cost and real-time performance", through a dynamic shift quantization modulation architecture, providing a brand-new technical path for external radiation source detection systems.
[0045] This application further achieves deep quantization noise suppression (SNR improvement ≥40dB), phase error optimization (≤0.8°), and dynamic range expansion (adapting to 86dB instantaneous fluctuations) by employing a phased perturbation injection and parameter linkage control mechanism, while maintaining low latency and low resource characteristics. It is applicable to high-precision signal processing scenarios such as external radiation source radar, 5G communication anti-interference, and medical ultrasound imaging.
[0046] Based on the foregoing embodiments, this application further includes controlling the start and stop of the first disturbance and the second disturbance in the following manner:
[0047] d1 perturbation is enabled when Amin < 1000 or k < 5;
[0048] When mn>8, the d2 perturbation is enabled;
[0049] When k≥12, and for a preset number of periods, all disturbances are turned off.
[0050] The generation of the first and second perturbations is implemented using a single-channel 32-bit LFSR. This single-channel 32-bit LFSR is configured with a specified polynomial to generate the perturbation sequences of the first and second perturbations in a time-division manner. For example, in some embodiments, the polynomial is... The perturbation sequences of the first perturbation d1 and the second perturbation d2 are generated in a time-division manner.
[0051] In some embodiments, the start-stop control of the first disturbance and the second disturbance is implemented using a finite state machine, which includes three states: IDLE, d1_ONLY, and d1+d2, with hysteresis tolerances of 2dB (first disturbance d1) and 3 bits (second disturbance d2).
[0052] In some embodiments, the dynamically calculated parameters include:
[0053] Right shift by bits: n = floor(log2(Amax / max(Amin,1)));
[0054] Signal-to-noise ratio: SNRdB = 20log10(Amax / Amin);
[0055] Adaptive shifter: when SNRdB ≥ 10dB, k = ceil(SNRdB / 3); when SNRdB < 10dB, k = 2;
[0056] Left shift bits: m = clamp(n + k, 3, B), where the overflow protection bits B = Bext - Bin.
[0057] In some embodiments, the sliding window has a window size of 2N, and a double-buffered parallel comparator tree is used to acquire Amax and Amin within a single cycle. In some embodiments, the overflow protection bit B is adapted to the corresponding bit input signal and is configurable; for example, B adapts to 8 / 16 / 24 / 32-bit input signals.
[0058] In some embodiments, the dynamic calculation is implemented based on a 32-bit barrel shifter, and the right shift number n and adaptive shift number k are calculated by lookup table or shift approximation, with a total delay ≤ 7 clock cycles.
[0059] In some examples, dynamic parameter computation involves a parallel comparator tree and a 32-bit barrel shifter, with a latency of ≤7 cycles @ 100MHz and a resource footprint of <500 LUTs.
[0060] In some examples, the root mean square error of the phase error of the subtraction separation is <1°, and the signal-to-noise ratio of weak signals is improved by ≥30dB.
[0061] The applicable scenario for the method in this application is radar: separation of strong direct waves (broadcast signals) and weak target echoes (dynamic range > 80dB) in external radiation source radar.
[0062] Communication: Suppress strong interference signals (such as interference from neighboring base stations) and extract weak user signals (signal-to-noise ratio <20dB).
[0063] Medical imaging: Separation of strong tissue reflection signals from weak blood flow signals in ultrasound imaging (signal-to-noise ratio <60dB).
[0064] Electronic reconnaissance: Separating strong enemy jamming from weak friendly communication signals in non-cooperative environments.
[0065] This application also proposes a dynamic shift quantization modulation strong and weak signal separation device for graded perturbation noise suppression, including:
[0066] The sign extension module is configured to extend the sign of the input signal;
[0067] The dynamic parameter calculation module uses a specified sliding window to statistically analyze the maximum value Amax and minimum value Amin of the symbol-extended signal and dynamically calculates the required parameters.
[0068] The shift amplification module amplifies the sign-extended signal by shifting it left by m bits based on the calculated parameters.
[0069] The phased perturbation injection module injects d1 / d2 perturbations after sign expansion and after left shift amplification, respectively.
[0070] The strong signal extraction module extracts strong signal components by shifting the signal right by n bits and left by n bits.
[0071] The weak signal separation module separates weak signals through subtraction.
[0072] The dynamic start-stop control module controls the disturbance injection state based on the parameters Amin, k, and mn.
[0073] The magnitude recovery module recovers the magnitude by shifting the strong and weak signal components to the right by m bits.
[0074] The output module uses symmetrical saturation truncation for strong signal components and asymmetrical truncation for weak signals.
[0075] This application aims to deeply integrate perturbation injection into the DSQM architecture. Through precise timing of staged perturbations and dynamic parameter linkage, it achieves synergistic optimization of noise suppression and hardware efficiency, realizing:
[0076] Quantization noise suppression: By injecting staged perturbations, the SNR of weak signals is increased from 25dB to over 40dB.
[0077] Phase fidelity enhancement: Reduces phase distortion caused by truncation error from 1.5° to below 0.8°.
[0078] Dynamic disturbance control: Construct a start-stop strategy with parameter linkage to adapt to instantaneous fluctuations in signal amplitude of 40-86dB.
[0079] Hardware efficiency is maintained: Based on the basic patent of 500 LUTs, the resource increment is <20% (total usage <600 LUTs).
[0080] This application proposes a phased perturbation-enhanced DSQM architecture, such as... Figure 2 As shown:
[0081] Signal flow: Input signal → sign extension → [d1 injection] → left shift amplification → [d2 injection] → dual-channel processing;
[0082] Strong signal path: Strong signal extraction → Strong signal post-processing (strong signal magnitude recovery → strong signal symmetrical truncation, restoring bit width output);
[0083] Weak signal path: Weak signal separation → Weak signal post-processing (weak signal magnitude recovery → weak signal asymmetric truncation, restoring bit width output);
[0084] Control flow: Symbol extended signal → Dynamic parameter calculation (after optimization) → Disturbance start / stop control → LFSR disturbance generation;
[0085] Key innovation modules:
[0086] Two-stage disturbance injection module:
[0087] d1 stage (after sign expansion): Breaks the periodicity of the original quantization noise, reducing the noise floor by 3-6 dB;
[0088] d2 stage (after left-shift amplification): Suppresses truncation error, reducing phase error of weak signals by 40%;
[0089] Dynamic amplitude formula:
[0090] |d1|=clamp(2^(mn), 1, Amin / 8)
[0091] |d2|=clamp(2^(mn), 1, (Amin< <m) / 8)
[0092] Bit-width mask design:
[0093] d1 mask: low log2(Amin) + 3 bits (covering 3σ fluctuations);
[0094] d2 mask: low m + log2(Amin) + 3 bits (adapted to bit width expansion after amplification).
[0095] A phased joint perturbation injection mechanism, the timing and physical significance of perturbation injection.
[0096] stage Injection site core role Quantization noise suppression mechanism d1 After symbol expansion, before left shift and magnification Breaking the periodicity of the original quantization noise improves the effective resolution of weak signals (ENOB + 0.5 bits). By using pseudo-random perturbation, the quantization noise energy is converted from discrete spectrum to white noise, reducing the noise floor by 3-6 dB. d2 After left-shifting and amplification, before strong signal extraction Suppressing quantization noise and truncation error after amplification The perturbation cancels out the low n-bit truncation error caused by the right-shift-left-shift operation, improving the linearity of weak signals (phase error reduced by 40%).
[0097] Precise design of disturbance amplitude and bit width
[0098] The formula for the magnitude of d1 is: |d1| = clamp(2^(mn), 1, Amin / 8)
[0099] Parameter linkage: 2^(mn): reflects the difference in dynamic range between strong and weak signals (m is the number of left shift bits, n is the number of strong signal truncation bits); ensures that the disturbance amplitude does not exceed the quantization noise suppression capability.
[0100] Amin / 8 constraint: to prevent disturbances from drowning out weak signals (Amin is the minimum signal value within the sliding window). Taking 8 facilitates shift calculation and ensures that the disturbance does not exceed 12.5% of the weak signal amplitude.
[0101] The lower limit covers the original quantization step size (Δ=1) to ensure effective noise whitening.
[0102] Physical meaning: Dynamic amplitude ensures the optimal perturbation ratio between the weak signal amplitude and the noise floor.
[0103] d1 bit width design: log2(Amin) + 3
[0104] log2(Amin): The original effective bit width of the weak signal (e.g., Amin = 328 → 9 bits).
[0105] +3 bits: Based on the 3σ principle, it covers statistical fluctuations and instantaneous errors.
[0106] The formula for the magnitude of d2 is: |d2| = clamp(2^(mn), 1, Amin2^m / 8)
[0107] Parameter linkage: Amin·2^m / 8: Matches the amplitude of the weak signal after amplification to prevent overflow (e.g., Amin=328, m=16 → upper limit of disturbance = 328×65536 / 8=2,687,872), where 2^m is the left shift amplification factor.
[0108] d2 bit width design: m + log2(Amin) + 3
[0109] m: Bit width expansion corresponding to the left shift amplification factor.
[0110] log2(Amin): Preserves the original details of weak signals.
[0111] +3 bits: Covers the noise floor fluctuation after amplification.
[0112] Dynamic disturbance start-stop control strategy
[0113] like Figure 3 As shown, the control logic is linked to the parameters (threshold is configurable).
[0114] Disturbance Activation conditions Discontinuation conditions Hysteresis tolerance Physical meaning d1 Amin < 1000 or k < 5 k ≥ 12 2dB Enabled when weak signals are close to the noise floor or the signal-to-noise ratio is insufficient; deactivated when the signal-to-noise ratio recovers. d2 mn>5 mn≤ 2 3 people Enabled when truncation error is significant, deactivated after one hour.
[0115] Parameter definition:
[0116] k=ceil(SNRdB / 3): Signal-to-noise ratio adaptive coefficient (SNR=20dB→k=7).
[0117] mn: Cut-off strength index (e.g., m=16, n=5→mn=11).
[0118] Hysteresis tolerance: 2dB (d1), 3 bits (d2) to prevent state oscillation.
[0119] State machine implementation
[0120] like Figure 3 As shown, the state is defined as follows:
[0121] S0 (IDLE): Disturbance off, monitoring parameters.
[0122] S1(d1_ONLY): Enable only the d1 perturbation.
[0123] S2(d1+d2): Jointly activate d1 and d2 disturbances.
[0124] State transition logic:
[0125] S0 → S1: When Amin < 1000 or k < 5;
[0126] S1 → S2: When mn>8 and lasts for 3 cycles;
[0127] S2 → S1: When mn ≤ 5;
[0128] S1 / S2 → S0: When k ≥ 12 and lasts for 5 periods
[0129] Hardware optimization implementation of this application embodiment:
[0130] Optimization of the dynamic parameter calculation module:
[0131] Quick calculation of log2(Amin):
[0132] Implementation method: Priority encoder + lookup table (LUT), input Amin, output the position of its most significant bit (e.g., 328→9).
[0133] Resource consumption: 16-bit input requires 4 levels of MUX, occupying approximately 25 LUTs.
[0134] Efficient generation of 2^(mn):
[0135] Implementation: Barrel shifter left shift (mn) bits, supports 32-bit shift (5-level MUX, 5ns delay).
[0136] Disturbance generation module (reuse optimization)
[0137] LFSR sharing: A single 32-bit LFSR generates a pseudo-random sequence and generates d1 and d2 perturbations in a time-division manner.
[0138] Polynomial: The cycle is -1.
[0139] Multiplexing logic: scaling the application for high and low bits separately
[0140] Resource reuse techniques
[0141] Shifter reuse: The barrel shifter of the dynamic parameter module is shared with the disturbance generation module to reduce redundant logic.
[0142] Dynamic mask generation: The bit-width mask is calculated in real time using the existing log2(Amin) and m value, avoiding the consumption of RAM due to pre-storage.
[0143] The embodiments of this application have the following advantages:
[0144] Quantization noise suppression: Through staged perturbation injection, the SNR of weak signals is improved from 25dB in the basic patent to more than 40dB.
[0145] Phase fidelity enhancement: The phase error caused by nonlinearity is reduced from 1.5° to below 0.9°, improving the accuracy of Doppler velocimetry and imaging.
[0146] Enhanced dynamic adaptability: Supports instantaneous fluctuations of up to 86dB in signal amplitude.
[0147] Hardware efficiency remains unchanged: resource usage increases by only 16% (500 → 580 LUTs), and real-time latency remains at 70ns@100MHz.
[0148] It should be noted that, in the embodiments of this application, 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 that element.
[0149] The sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.
[0150] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.
[0151] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims. All of these forms are within the protection scope of this application.
Claims
1. A method for separating strong and weak signals in dynamic shift quantization modulation with graded perturbation noise suppression, characterized in that, include: Sign extension of the input signal; For the symbol-extended signal, the maximum value Amax and minimum value Amin of the signal are statistically analyzed using a specified sliding window, and the required parameters are dynamically calculated. For the symbol-spread signal, a first perturbation d1 is injected, and according to the calculated parameters, the symbol-spread signal injected with the first perturbation d1 is shifted left by m bits to amplify it, and a second perturbation d2 is injected into the amplified signal. For the signal injected with the second disturbance d2, the strong signal component is extracted by shifting it right by n bits and left by n bits in sequence, and the weak signal is separated by subtraction. The magnitude is recovered by shifting strong and weak signal components to the right by m bits. Strong signal components are symmetrically saturated and truncated, while weak signals are output with asymmetrical truncation. The first perturbation d1 has an amplitude of |d1|=clamp(2^(mn),1,Amin / 8) and a bit width of low log2(Amin)+3 bits; The second perturbation d2 has an amplitude of |d2|=clamp(2^(mn),1,Amin·2^m / 8) and a bit width of low m+log2(Amin)+3 bits; Sign extension of the input signal includes increasing the bit width from B... in Extend to B ext =2×B in High-order fill sign bit; The parameters calculated dynamically include: Right shift by bits: n = floor(log2(Amax / max(Amin,1))); Signal-to-noise ratio: SNRdB = 20log10(Amax / Amin); Adaptive shift value k: when SNRdB≥10dB, k=ceil(SNRdB / 3); when SNRdB<10dB, k=2; Left shift bits: m = clamp(n + k, 3, B), where B = Bext - Bin, and B is the overflow protection bits.
2. The method for separating strong and weak signals by dynamic shift quantization modulation with graded perturbation noise suppression as described in claim 1, characterized in that, It also includes controlling the start and stop of the first and second disturbances in the following ways: d1 perturbation is enabled when Amin < 1000 or k < 5; When mn>8, the d2 perturbation is enabled; When k≥12, and for a preset number of periods, all disturbances are turned off; The generation of the first and second perturbations is achieved using a single-channel 32-bit LFSR, which is configured with a specified polynomial to generate the perturbation sequences of the first and second perturbations in a time-division manner.
3. The method for separating strong and weak signals by dynamic shift quantization modulation with graded perturbation noise suppression as described in claim 2, characterized in that, The start-stop control of the first and second disturbances is implemented using a finite state machine, which includes three states: IDLE, d1_ONLY, and d1+d2, with a hysteresis tolerance of 2dB and 3 bits.
4. The method for separating strong and weak signals by dynamic shift quantization modulation with graded perturbation noise suppression as described in claim 1, characterized in that, The sliding window has a window size of 2N, and a double-buffered parallel comparator tree is used to obtain Amax and Amin in a single cycle.
5. The method for separating strong and weak signals by dynamic shift quantization modulation with graded perturbation noise suppression as described in claim 1, characterized in that, The dynamic calculation is based on a 32-bit barrel shifter. The right shift number n and the adaptive shift number k are calculated by lookup table or shift approximation, with a total delay of ≤7 clock cycles.
6. A dynamic shift quantization modulation strong / weak signal separation device for graded perturbation noise suppression, characterized in that, include: The sign extension module is configured to extend the sign of the input signal; The dynamic parameter calculation module uses a specified sliding window to statistically analyze the maximum value Amax and minimum value Amin of the symbol-extended signal and dynamically calculates the required parameters. The shift amplification module amplifies the sign-extended signal by shifting it left by m bits based on the calculated parameters. The staged perturbation injection module is used to inject a first perturbation d1 and a second perturbation d2 into the signal before and after the sign-extended signal is amplified by shifting it left by m bits. The strong signal extraction module extracts the strong signal components from the signal injected with the second disturbance d2 by shifting it right by n bits and left by n bits. The weak signal separation module separates weak signals through subtraction. The magnitude recovery module right-shifts strong and weak signal components by m bits to recover the magnitude. The output module uses symmetrical saturation truncation for strong signal components and asymmetrical truncation for weak signals. The first perturbation d1 has an amplitude of |d1|=clamp(2^(mn),1,Amin / 8) and a bit width of low log2(Amin)+3 bits; The second perturbation d2 has an amplitude of |d2|=clamp(2^(mn),1,Amin·2^m / 8) and a bit width of low m+log2(Amin)+3 bits; Sign extension of the input signal includes increasing the bit width from B... in Extend to B ext =2×B in High-order fill sign bit; The parameters calculated dynamically include: Right shift by bits: n = floor(log2(Amax / max(Amin,1))); Signal-to-noise ratio: SNRdB = 20log10(Amax / Amin); Adaptive shift value k: when SNRdB≥10dB, k=ceil(SNRdB / 3); when SNRdB<10dB, k=2; Left shift bits: m = clamp(n + k, 3, B), where B = Bext - Bin, and B is the overflow protection bits.