A shortwave broadband radio signal receiving and channelization processing system

Abstract

This invention relates to the fields of radio communication and digital signal processing technology, specifically to a shortwave broadband radio signal receiving and channelization processing system, comprising: a broadband digital receiving front-end for receiving analog radio frequency signals in the shortwave band and converting them into high-speed digital signals; a spectrum analysis and signal detection module for performing spectrum analysis on the high-speed digital signals, identifying and locating the various signal sources contained therein; a dynamic channelization resource management module for calculating and allocating corresponding channelization processing resources in real time; and a reconfigurable parallel channelization processing engine for performing parallel dynamic digital down-conversion and filtering extraction on the high-speed digital signals, thereby extracting multiple independent narrowband signal data streams; capable of adaptively detecting and simultaneously receiving multiple independent signals within the broadband spectrum; this invention effectively solves the frequency offset problem caused by the Doppler effect or medium changes, reducing the complexity of back-end signal processing.

Description

Technical Field

[0001] This invention relates to the field of radio communication and digital signal processing technology, specifically to a system for receiving and channelizing shortwave broadband radio signals. Background Technology

[0002] With the continuous evolution of radio communication technology, the shortwave communication environment is becoming increasingly complex, and spectrum resources are exhibiting dynamic and highly challenging characteristics. In particular, in scenarios such as ocean communication, aviation monitoring, and emergency command, shortwave band signals generally have physical characteristics such as low density, large frequency drift, and irregular signal bandwidth.

[0003] Currently, the reception and processing of shortwave broadband signals generally employs a traditional fixed-grid channelization architecture. This architecture is typically based on polyphase filter bank technology, which pre-divides the broadband spectrum into several sub-channels with fixed bandwidths and center frequencies. The system allocates hardware resources according to preset fixed parameters, performing uniform channelization processing and energy detection across the entire frequency band. However, this traditional fixed-grid processing method has significant limitations when dealing with non-steady-state shortwave signals. Due to the fixed parameters of the processing channels, when the signal crosses the channel boundary, cross-channel splicing processing leads to significant phase distortion and amplitude collapse. Simultaneously, the fixed filter bandwidth is difficult to adapt to the varying actual bandwidth of the signal, easily causing truncation distortion at the edges of the broadband signal and introducing unnecessary out-of-band noise when processing narrowband signals, reducing the signal-to-noise ratio. Furthermore, indiscriminate resource allocation across all frequency bands results in a significant waste of computing power in idle frequency bands with no signal when the signal is sparse, failing to achieve efficient utilization of hardware resources. Therefore, how to break the limitations of the fixed grid and achieve adaptive high-fidelity reception and dynamic resource scheduling for shortwave broadband signals has become an urgent problem to be solved in this field. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a system for receiving and channelizing shortwave broadband radio signals. Specifically, the technical solution of this invention includes:

[0005] A broadband digital receiver front end is used to receive analog radio frequency signals in the shortwave band and convert them into high-speed digital signals;

[0006] A spectrum analysis and signal detection module is connected to the broadband digital receiving front end and is used to perform spectrum analysis on the high-speed digital signal to identify and locate the various signal sources contained therein.

[0007] The dynamic channelization resource management module is connected to the spectrum analysis and signal detection module and is used to calculate and allocate the corresponding channelization processing resources in real time based on the identified signal source parameters.

[0008] A reconfigurable parallel channelization processing engine is connected to the broadband digital receiving front end and the dynamic channelization resource management module, respectively. It is used to perform parallel dynamic digital downconversion and filtering extraction on the high-speed digital signal according to the allocated resource parameters, thereby extracting multiple independent narrowband signal data streams.

[0009] The system constitutes a complete shortwave broadband receiver, capable of adaptively detecting and simultaneously receiving multiple independent signals within the broadband spectrum.

[0010] Preferably, the spectrum analysis and signal detection module includes:

[0011] A spectrum calculation unit is used to perform time-frequency transformation on the high-speed digital signal to obtain its short-time spectrum distribution;

[0012] The signal identification unit is used to analyze the short-time spectral distribution and, based on the characteristics of spectral entropy or energy concentration, distinguish between potential signal regions with high signal-to-noise ratios and noise background regions.

[0013] Preferably, the signal recognition unit determines the center position and boundary of each potential signal region in the frequency domain through gradient search or contour extraction algorithms.

[0014] Preferably, the dynamic channelization resource management module calculates the digital down-conversion local oscillator frequency and filter bandwidth parameters corresponding to each signal based on the signal center position and boundary output by the signal identification unit, and sends them to the reconfigurable parallel channelization processing engine.

[0015] Preferably, the reconfigurable parallel channelized processing engine includes multiple parallel processing channels, each channel containing:

[0016] A digital downconverter, the local oscillator frequency of which is dynamically configured by the dynamic channelization resource management module;

[0017] A variable bandwidth filter bank, whose filter coefficients and decimation factors are generated in real time by the dynamic channelization resource management module based on the allocated bandwidth parameters.

[0018] Preferably, the spectrum analysis and signal detection module works continuously. When a drift in the center frequency of the tracked signal is detected, the dynamic channelization resource management module updates the local oscillator frequency parameters of the corresponding channel in real time to achieve tracking compensation for the signal frequency drift.

[0019] Preferably, the spectrum analysis and signal detection module and the dynamic channelization resource management module are integrated into a field-programmable gate array (FPGA), which utilizes its internal storage and arithmetic units to achieve high-speed parallel processing.

[0020] Preferred options also include:

[0021] The spectrum occupancy assessment module is used to calculate the sparsity of signal activity within a wideband spectrum.

[0022] The mode switching controller is used to enable the dynamic channelization resource management module when the spectrum is sparse, and switch to a fixed full-band scanning reception mode when the spectrum is dense.

[0023] Compared with the prior art, the present invention has the following beneficial effects:

[0024] 1. This invention constructs an adaptive fluid receiver architecture based on energy potential, which abandons the traditional fixed grid limitation and allows the center frequency of the processing channel to change continuously with the signal position. This mechanism effectively eliminates the phase distortion and amplitude collapse caused by cross-channel splicing in the traditional polyphase filter architecture, realizes high-fidelity reception of signals at any position in the broadband spectrum, and significantly improves the phase consistency of signal extraction.

[0025] 2. This invention introduces spectral entropy potential field mapping and gradient search algorithm to realize liquid adaptive segmentation of signal boundaries; this method can effectively penetrate strong noise interference, accurately identify the signal center and boundary, and dynamically generate matching filter parameters accordingly; this on-demand configuration not only prevents the truncation distortion of broadband signals by fixed bandwidth, but also avoids the introduction of unnecessary out-of-band noise in narrowband signal processing, and maximizes the optimization of the signal-to-noise ratio of the demodulated output.

[0026] 3. This invention establishes a closed-loop automatic frequency control mechanism with dynamic engagement capability for non-steady-state signals; in response to the slow frequency drift characteristics caused by ionospheric instability in shortwave channels, the system can monitor and update the local oscillator frequency parameters in real time for tracking compensation; this ensures the phase continuity of the demodulated baseband data, effectively solves the frequency offset problem caused by Doppler effect or medium changes, and reduces the complexity of back-end signal processing;

[0027] 4. This invention realizes signal-driven dynamic resource scheduling and intelligent mode switching; in signal-sparse scenarios, hardware resources are allocated only for effective signals, eliminating the wasted computing power of idle frequency bands and significantly reducing system power consumption; at the same time, through the dual-mode switching strategy, it automatically reverts to full-scan mode when the spectrum is congested, preventing system paralysis due to resource exhaustion and ensuring monitoring stability in extreme electromagnetic environments. Attached Figure Description

[0028] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0029] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0031] Example 1:

[0032] Please see Figure 1 A system for receiving and channelizing shortwave broadband radio signals, comprising:

[0033] A broadband digital receiver front end is used to receive analog radio frequency signals in the shortwave band and convert them into high-speed digital signals;

[0034] A spectrum analysis and signal detection module is connected to the broadband digital receiving front end and is used to perform spectrum analysis on the high-speed digital signal to identify and locate the various signal sources contained therein.

[0035] The dynamic channelization resource management module is connected to the spectrum analysis and signal detection module and is used to calculate and allocate the corresponding channelization processing resources in real time based on the identified signal source parameters.

[0036] A reconfigurable parallel channelization processing engine is connected to the broadband digital receiving front end and the dynamic channelization resource management module, respectively. It is used to perform parallel dynamic digital downconversion and filtering extraction on the high-speed digital signal according to the allocated resource parameters, thereby extracting multiple independent narrowband signal data streams.

[0037] The system constitutes a complete shortwave broadband receiver, capable of adaptively detecting and simultaneously receiving multiple independent signals within the broadband spectrum.

[0038] This embodiment provides a shortwave broadband radio signal receiving and channelization processing system. Addressing the characteristics of low signal density but large frequency drift and irregular bandwidth in the shortwave band, this system abandons the traditional fixed-grid channelization architecture and adopts an adaptive fluid-like receiving architecture based on energy potential. Here, the custom terms and abbreviations used in this specification are clearly defined and uniformly explained: Energy potential is defined as a scalar field mapping value whose value is positively correlated with the reciprocal of the local spectral entropy, physically representing the probability density of the signal's existence in the frequency domain; Fluid-like receiving architecture is defined as a non-fixed-grid channelization resource allocation method, whose processing channel's center frequency can arbitrarily and continuously change with the signal position, adapting to the container, i.e., the signal spectrum, like a fluid; Spectral entropy potential field mapping is defined as the mathematical process of converting a one-dimensional spectral amplitude sequence into a two-dimensional potential energy topology; Furthermore, the spectrum analysis and signal detection module mentioned in this document may be simply referred to as the detection module or signal detection module in the following text; The reconfigurable parallel channelization processing engine may be simply referred to as the processing engine or channelization processing engine.

[0039] In this embodiment, the overall system workflow is as follows: a broadband digital receiving front-end conditions the analog radio frequency signal sensed by the antenna; this front-end includes a high dynamic range analog-to-digital converter (ADC), whose sampling rate is set to... For example, 80Msps directly performs bandpass sampling on the entire shortwave frequency band from 1.5MHz to 30MHz, outputting a high-speed parallel digital signal stream. The spectrum analysis and signal detection module serves as the panoramic perception center of the system. Its core is no longer simple power spectrum detection, but introduces a spectral entropy potential field mapping mechanism. It calculates the orderliness of the broadband spectrum in real time, transforming the signal recognition problem into a topological analysis problem of the thermodynamic potential energy field, thereby accurately locating the center position of each signal source.

[0040] The dynamic channelization resource management module, acting as the system's scheduling brain, dynamically calculates the required demodulation resources based on the potential energy field topology output by the detection module. Unlike traditional systems that allocate fixed channel numbers, this module calculates precise physical frequency and bandwidth parameters and maps them to hardware resource configuration instructions. The reconfigurable parallel channelization processing engine serves as the execution mechanism, containing... For example, 32 completely independent digital down-conversion (DDC) channels; these channels are no longer limited to a fixed channel grid but are in a floating state; each channel, according to assigned parameters, precisely targets the energy center of the signal like a straw to extract and filter, outputting... Independent narrowband baseband signal ;

[0041] This embodiment achieves on-demand reception of shortwave broadband signals by constructing an adaptive fluid receiver architecture based on energy potential. In scenarios such as ocean-going ship communications where signals are sparse but frequencies are variable, this system breaks through the bandwidth limitations of fixed grid channelization and effectively solves the phase distortion problem caused by cross-channel splicing in the traditional polyphase filter bank (PFB) architecture. A hardware test platform based on Kintex-7 FPGA was built. Experimental results show that when 32 mixed test signals with different bandwidths (3kHz to 20kHz) are input, the phase consistency error of the narrowband signal extracted by this system is less than 1.5 degrees. Compared with the traditional PFB architecture, it eliminates the amplitude collapse phenomenon at cross-channel splicing. In a sparse environment with a spectrum occupancy rate of 10%, the dynamic resource allocation mechanism reduces the overall power consumption of the system by about 45% compared to the fully enabled mode.

[0042] Example 2:

[0043] The spectrum analysis and signal detection module includes:

[0044] A spectrum calculation unit is used to perform time-frequency transformation on the high-speed digital signal to obtain its short-time spectrum distribution;

[0045] The signal identification unit is used to analyze the short-time spectral distribution and, based on the characteristics of spectral entropy or energy concentration, distinguish between potential signal regions with high signal-to-noise ratios and noise background regions.

[0046] This embodiment details the implementation of the spectrum analysis and signal detection module, focusing on introducing spectral entropy as the criterion for determining the presence of a signal; the spectrum calculation unit processes the input high-speed digital signal. A windowed overlapping Fast Fourier Transform (FFT) is performed; in this embodiment, a method is used. Windowing to suppress spectral leakage, FFT points Set to 4096 points, overlap rate to 50%; output is instantaneous spectral amplitude sequence. ,in It is a frequency index, and In order to distinguish signals in a strong noise background, the signal recognition unit defines a local spectral entropy density. Unlike energy detection, which relies solely on amplitude, spectral entropy utilizes the ordered nature of signals in the frequency domain (i.e., concentrated energy) and the disordered nature of noise (i.e., randomly distributed energy). The local spectral entropy density calculation model is as follows:

[0047] ;

[0048] in, The source is generated in real time by the computing unit, and its physical meaning is frequency index. The local spectral entropy density at a given location, in dimensionless units;

[0049] The source is a system preset, and the physical meaning is based on an index. A set of local sliding windows centered on the object, for example, taking ;

[0050] The source is the spectrum calculation unit, and the physical meaning is the first [unit] within the window. The spectral amplitude at each frequency point is expressed in volts.

[0051] The source is the spectrum calculation unit, and the physical meaning is the first [unit] within the window. The spectral amplitude at each frequency point, in volts.

[0052] The source is calculated, and the physical meaning is the first [unit] within the window. Normalized power percentage of each frequency point, in dimensionless units;

[0053] The source is calculated based on system parameters, and the calculation formula is:

[0054] ;

[0055] in, Minimum signal bandwidth, For FFT points, The sampling rate, physically represented by the sliding window radius, is expressed in points. Based on this, traditional energy detection is prone to false alarms at low signal-to-noise ratios. By introducing the above formula, when the window... When there is a single-tone or narrowband signal in memory, the energy is concentrated at a few frequency points. The distribution is extremely uneven, resulting in high entropy. A significant decrease, i.e., approaching 0, forms a potential well; conversely, for white noise, the energy is uniformly distributed, and the entropy value... A higher value indicates the formation of a potential barrier. Furthermore, this embodiment also provides an alternative detection scheme based on the peak power spectrum. When the system operates in a high signal-to-noise ratio environment, it can switch to the energy concentration detection mode. The energy concentration calculation model is as follows:

[0056] ;

[0057] in, The source is calculated, and the physical meaning is frequency index. The ratio of energy concentration at a given location, in dimensionless units;

[0058] The source is calculated, and its physical meaning is the average power spectral density estimate within a local window, with units of volts².

[0059] The source is a system preset, and its physical meaning is a window. The total number of frequency points included, i.e. The unit is a constant; it should be noted that in the above formula... and The calculation implicitly assumes a normalized impedance, that is, the system load impedance is set. And it ignores frequency resolution bandwidth. The influence of dimensions; due to Since the ratio of physical quantities of the same dimension is equal, the actual impedance value and the bandwidth constant cancel each other out in the numerator and denominator. Therefore, the square of the spectral amplitude, i.e., volts², is directly used to characterize the relative power for calculation, which will not affect the evaluation result of energy concentration. It should be understood that the above formula uses the square of the amplitude to characterize the power only as one implementation method. All calculation models that characterize the degree of energy concentration in the frequency domain based on the statistical characteristics of the spectral amplitude are within the protection scope of this invention.

[0060] In this mode, when Greater than the preset peak threshold When the signal is strong, it indicates that there is a narrowband signal with highly concentrated energy in the region. This method has a small computational load and is suitable for the rapid capture of strong burst signals. The signal identification unit constructs a full-band distribution map based on the selected pattern, i.e., spectral entropy or energy concentration, and marks the regions whose feature values ​​meet the threshold conditions as potential signal regions with high signal-to-noise ratio.

[0061] This embodiment establishes a flexible signal detection architecture by introducing two complementary decision mechanisms: local spectral entropy density and energy concentration. In urban radio monitoring scenarios with complex electromagnetic environments, the spectral entropy mechanism can effectively penetrate the interference of strong impulse noise, while the energy concentration mechanism can quickly lock onto strong signals. The combination of the two greatly improves the system's probability of detecting various signals and its adaptability.

[0062] Example 3:

[0063] The signal recognition unit determines the center position and boundary of each potential signal region in the frequency domain through gradient search or contour extraction algorithms.

[0064] This embodiment further refines the region delineation method in the signal recognition unit, employing a gradient search algorithm or contour extraction algorithm to accurately cut the signal boundary; after constructing the spectral entropy distribution map, we regard the spectrum as a terrain, where low-entropy regions correspond to valleys, i.e., signals, and high-entropy regions correspond to peaks, i.e., noise;

[0065] The gradient search algorithm performs the extreme point search step; the system traverses the spectral entropy values ​​across the entire frequency band. Finding local minimum points ,satisfy and ; this point The spectral kernel centers are marked as potential signals; a gradient search, i.e., a gradient ramping step, is performed; the system moves from the spectral kernel centers... Perform gradient search to the left and right respectively; search to the left: find the index. , making Follow Decrease and monotonically increase until... or This means the gradient is flat, indicating a noisy plateau; search to the right: similarly find the index. ;

[0066] In the above algorithm, the key threshold is set based on the following: spectral entropy threshold. It is determined based on the system initialization calibration, that is Here, and The acquisition method is as follows: after the system is powered on, the noise floor calibration mode is executed, and data is collected in the absence of RF input. For example, 4096 points of spectral entropy data, calculate its mean. and standard deviation And store; gradient threshold Set to 1.5 times the noise plateau volatility, i.e. ,in, Defined as the gradient spectrum of the above calibration data, i.e., the first-order difference, and its standard deviation; edge threshold. , set as This is to distinguish the edges of the effective signal from the random fluctuations of noise;

[0067] As another parallel implementation method mentioned in the embodiments, the contour extraction algorithm has the following specific implementation steps:

[0068] Edge enhancement: for spectral entropy sequences Perform a first-order difference operation to obtain the gradient spectrum. ;

[0069] Edge detection: Set positive and negative edge thresholds ;when The time marker is the falling edge, i.e., the candidate for the left boundary of the signal. The time marker is marked as the rising edge, i.e., a candidate for the right boundary of the signal;

[0070] Contour pairing: Search for adjacent falling edge and rising edge pairs across the entire frequency band. If the distance between them meets the minimum bandwidth limit, it is confirmed as a valid signal contour, and its center position and boundary are determined by this edge pair; perform the region locking step; the system determines the effective coverage area of ​​the signal source in the frequency domain as follows. ;

[0071] This embodiment utilizes gradient search or contour extraction algorithms to achieve liquid adaptive segmentation of signal boundaries. In emergency communication command scenarios that process mixed-mode signals, the algorithm can automatically match the actual physical bandwidth of OFDM broadband signals and CW narrowband signals. This not only prevents the truncation distortion of broadband signal edges caused by fixed filters, but also avoids the introduction of unnecessary out-of-band noise in narrowband signal processing, thus optimizing the signal-to-noise ratio of subsequent demodulation.

[0072] Example 4:

[0073] The dynamic channelization resource management module calculates the digital down-conversion local oscillator frequency and filter bandwidth parameters corresponding to each signal based on the signal center position and boundary output by the signal identification unit, and sends them to the reconfigurable parallel channelization processing engine.

[0074] This embodiment details how the dynamic channelization resource management module converts the identified boundary parameters into hardware control parameters; this module receives the boundary index output by the signal identification unit. and The physical parameters are calculated using the following logic, and the parameter calculation model is as follows:

[0075] Here, the frequency resolution is clearly defined. The calculation method is as follows:

[0076] ;

[0077] in, Sampling rate, The number of FFT points;

[0078] ;

[0079] in, Down-conversion local oscillator frequency, Signal boundary index;

[0080] ;

[0081] in, : Filter cutoff bandwidth : Protection bandwidth factor; Protection bandwidth factor The value is set to 1.2, determined based on filter transition band design principles. This aims to reserve a 20% bandwidth margin to accommodate the filter's roll-off characteristics and prevent amplitude attenuation at the passband edge. Based on this, resource allocation logic is executed; the module maintains a resource idle pool; when a new... When the signal disappears, i.e. the spectral entropy rises again, the module requests an idle DDC channel ID from the pool and sends the above parameters to the channel via the control bus; if the signal disappears, i.e. the spectral entropy rises again, the module reclaims the channel ID.

[0082] This embodiment constructs a signal-driven dynamic resource scheduling mechanism, realizing real-time mapping of physical parameters to hardware instructions. In the aviation band monitoring scenario where spectrum resources are scarce, this module only allocates DDC channels for actual existing signals. This on-demand allocation strategy fundamentally eliminates the wasted computing power of idle frequency bands, enabling limited hardware resources to maximize their service to the acquisition of effective signals.

[0083] Example 5:

[0084] The reconfigurable parallel channelized processing engine includes multiple parallel processing channels, each containing:

[0085] A digital downconverter, the local oscillator frequency of which is dynamically configured by the dynamic channelization resource management module;

[0086] A variable bandwidth filter bank, whose filter coefficients and decimation factors are generated in real time by the dynamic channelization resource management module based on the allocated bandwidth parameters.

[0087] This embodiment details the internal architecture of the reconfigurable parallel channelization processing engine; each processing channel is completely isomorphic and includes the following key components: a digital downconverter (DDC) as the core component, comprising a numerically controlled oscillator (NCO) and a complex multiplier; the NCO operates based on the input local oscillator frequency parameters. Generating sine and cosine waveforms using the CORDIC algorithm ; Input high-speed signal Multiplying this waveform shifts the center frequency spectrum of the target signal to zero intermediate frequency; the variable bandwidth filter bank adapts to dynamically changing bandwidth. It employs a heterogeneous cascaded structure of a cascaded integrator comb filter (CIC) and a programmable finite impulse response (FIR) filter; specifically, the CIC filter is responsible for high-order decimation, and its decimation factor... Calculated dynamically based on the following model:

[0088] ;

[0089] in, Extraction multiplier; Sampling rate; Signal bandwidth; Oversampling factor; Oversampling factor The setting logic is based on the Nyquist sampling theorem and the filter transition band design principle: The larger the value, the better the passband flatness of the CIC filter, but the greater the processing burden on the subsequent FIR filter. The specific values ​​can be flexibly adjusted according to the system's specific requirements for stopband attenuation and passband ripple. In the preferred embodiment, as verified by Matlab system-level simulation, when When the value is between 2.5 and 3.0, it can maintain a passband ripple of less than 0.1 dB while keeping the FIR order within the limits allowed by hardware resources; the FIR filter is used for pass shaping and final bandwidth limiting; its filter coefficients It is not fixed, but rather generated in real time by the resource management module based on a pre-stored normalized prototype filter coefficient table; this coefficient table pre-stores a set of... Window weighted Function coefficients Table length The value is 1024; it should be noted that this embodiment uses linear interpolation algorithm as a preferred method, but in practical applications, the real-time generation is not limited to linear interpolation, and can also be achieved using mathematical methods such as polynomial fitting, spline interpolation, or direct table lookup; real-time generation of target coefficients When using linear interpolation, the specific calculation is as follows:

[0090] ;

[0091] in, Target coefficient Prototype coefficients

[0092] : Interpolation of the decimal part, Integer index, Interpolation step ratio; parameter Defined as The effective order of the target filter Based on the following functional relationship Dynamically determined:

[0093] ;

[0094] in, : Target order Window function constant, Sampling rate Signal bandwidth : Transition zone coefficient; The value is 5.5, corresponding to Window characteristics The value is set to 0.15, meaning the transition band width is set to 15% of the signal bandwidth; using the above algorithm, the module will calculate... Write coefficient ;

[0095] This embodiment employs a heterogeneous filtering architecture cascaded with CIC and reconfigurable FIR, combined with linear interpolation coefficient generation technology based on prototype tables, to achieve stepless and continuously adjustable channel bandwidth. Resource consumption and performance comparison tests were conducted on the reconfigurable parallel channelization processing engine proposed in this embodiment. Test data shows that when processing 32 parallel signals, compared to the traditional DDC scheme based on frequency sampling, the DSP slice resource utilization rate of this engine is reduced by 28%. Furthermore, due to the adoption of dynamic coefficient generation technology, while supporting continuously variable bandwidth from 500Hz to 50kHz, the stopband attenuation remains consistently above 60dB, verifying the engine's advantage in balancing low power consumption and high performance.

[0096] Example 6:

[0097] The spectrum analysis and signal detection module works continuously. When a drift in the center frequency of the tracked signal is detected, the dynamic channelization resource management module updates the local oscillator frequency parameters of the corresponding channel in real time to achieve tracking compensation for the signal frequency drift.

[0098] This embodiment describes the system's frequency drift tracking compensation mechanism. In shortwave communication, due to the instability of the ionosphere, the signal frequency often experiences slow drift. The spectrum analysis and signal detection module continuously monitors the spectral core center of the locked signal during continuous operation. The system executes tracking and compensation logic; a drift tolerance threshold is set. For example, 50Hz; in response to detecting the current moment Spectral core center frequency Compared to the previous moment Recorded center frequency The absolute value of the difference exceeds At that time, that is ,in, This is the frequency drift. The drift tolerance threshold triggers an update mechanism; a real-time update step is executed; the dynamic channelization resource management module immediately calculates the new local oscillator frequency. Based on the bit width of the NCO accumulator In this embodiment, a 32-bit phase increment control word is used for calculation. :

[0099] ;

[0100] The calculated integer pass - The bus writes to the phase increment register of the corresponding DDC channel in real time; due to the phase continuity design of the NCO, this frequency switching is smooth and will not cause phase abrupt changes or packet loss in the demodulated data.

[0101] This embodiment establishes a closed-loop automatic frequency control mechanism, which gives the system the ability to dynamically engage with non-steady-state signals. In long-distance shortwave communication scenarios with severe ionospheric disturbances, this mechanism can compensate for the slow drift of the center frequency caused by the Doppler effect or changes in the medium in real time, ensuring the phase continuity of the demodulated baseband data and effectively reducing the complexity requirements of the back-end signal processing algorithm for frequency offset estimation.

[0102] Example 7:

[0103] The spectrum analysis and signal detection module and the dynamic channelization resource management module are integrated into a field-programmable gate array (FPGA), which utilizes its internal storage and arithmetic units to achieve high-speed parallel processing.

[0104] This embodiment relates to the hardware integration and acceleration scheme of the system; the entire system is integrated into a high-performance field-programmable gate array (FPGA), such as a Xilinx Kintex-7 or Virtex-7 series; regarding memory utilization, the internal memory of the FPGA is utilized. abbreviation A dual-port memory for the spectral entropy potential energy map is constructed to achieve pipelined parallelism of spectrum writing and gradient search algorithm reading. In terms of computing unit utilization, DSP48E slices are used to implement butterfly operation of FFT, logarithmic approximation operation in spectral entropy calculation (which is implemented by polynomial fitting), and multiply-add operation in DDC. In terms of parallel design, 32 channelized processing channels are logically parallelized and physically share part of the DSP resources through time-division multiplexing (TDM) to balance speed and area.

[0105] This embodiment fully utilizes the internal features of the FPGA. The parallel architecture with DSP slicing constructs a fully hardware pipeline for signal processing. In tactical radio reconnaissance scenarios requiring millisecond-level response, this integrated solution compresses the end-to-end delay of broadband spectrum analysis and multi-channel channelization to the microsecond level, ensuring that the system can capture transient frequency hopping signals and meeting the extremely high real-time signal interception requirements.

[0106] Example 8:

[0107] This system also includes:

[0108] The spectrum occupancy assessment module is used to calculate the sparsity of signal activity within a wideband spectrum.

[0109] The mode switching controller is used to enable the dynamic channelization resource management module when the spectrum is sparse, and switch to a fixed full-band scanning reception mode when the spectrum is dense.

[0110] This embodiment introduces an intelligent mode switching mechanism to cope with extreme spectrum environments; the spectrum occupancy assessment module periodically calculates the total bandwidth ratio of potential high signal-to-noise ratio signal regions across the entire frequency band, which is defined as the spectrum occupancy rate. The mode switching controller sets a sparsity threshold. For example, 30%; in response to The system enters dynamic mode, i.e., the default mode; in this mode, the spectrum is considered sparse, and the system activates the aforementioned dynamic channelization resource management module to allocate resources on demand, achieving low-power, high-precision adaptive reception; in response to The system enters full scan mode, i.e., back-off mode; this indicates that the spectrum is extremely congested, such as during competition or strong interference, at which point the potential energy field may not be able to form an effective independent potential well; the controller automatically switches strategies, bypassing the dynamic management module, and... Each processing channel is configured as a scanning receiver with a fixed frequency interval, or polls and monitors key frequency bands sequentially.

[0111] This embodiment designs a dual-mode switching strategy based on spectral sparsity, which solves the performance bottleneck of the adaptive algorithm in extreme congestion environments. In the scenario of international shortwave broadcasting bands with fierce spectrum competition, this mechanism ensures that the system can smoothly fall back to full scan mode when the signal is overloaded, preventing system paralysis due to resource exhaustion and ensuring that basic spectrum monitoring capabilities can be maintained in any electromagnetic environment.

[0112] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A system for receiving and channelizing shortwave broadband radio signals, characterized in that, include: A broadband digital receiver front end is used to receive analog radio frequency signals in the shortwave band and convert them into high-speed digital signals; A spectrum analysis and signal detection module is connected to the broadband digital receiving front end and is used to perform spectrum analysis on the high-speed digital signal to identify and locate the various signal sources contained therein. The dynamic channelization resource management module is connected to the spectrum analysis and signal detection module and is used to calculate and allocate the corresponding channelization processing resources in real time based on the identified signal source parameters. A reconfigurable parallel channelization processing engine is connected to the broadband digital receiving front end and the dynamic channelization resource management module, respectively. It is used to perform parallel dynamic digital downconversion and filtering extraction on the high-speed digital signal according to the allocated resource parameters, thereby extracting multiple independent narrowband signal data streams. The system constitutes a complete shortwave broadband receiver, capable of adaptively detecting and simultaneously receiving multiple independent signals within the broadband spectrum; The spectrum analysis and signal detection module includes: A spectrum calculation unit is used to perform time-frequency transformation on the high-speed digital signal to obtain its short-time spectrum distribution; The signal recognition unit is used to analyze the short-time spectral distribution and distinguish between potential signal regions with high signal-to-noise ratios and noise background regions based on the characteristics of spectral entropy or energy concentration. The signal recognition unit determines the center position and boundary of each potential signal region in the frequency domain through gradient search or contour extraction algorithms. The dynamic channelization resource management module calculates the digital down-conversion local oscillator frequency and filter bandwidth parameters corresponding to each signal based on the signal center position and boundary output by the signal identification unit, and sends them to the reconfigurable parallel channelization processing engine. The local spectral entropy density calculation model is as follows: ； in, The source is generated in real time by the computing unit, and its physical meaning is frequency index. The local spectral entropy density at a given location, in dimensionless units; The source is a system preset, and the physical meaning is frequency indexing. A set of local sliding windows centered on the target; The source is the spectrum calculation unit, and the physical meaning is the first [unit] within the window. The spectral amplitude at each frequency point is expressed in volts. The source is the spectrum calculation unit, and the physical meaning is the first [unit] within the window. The spectral amplitude at each frequency point, in volts. The source is calculated, and the physical meaning is the first [unit] within the window. Normalized power percentage of each frequency point, in dimensionless units; The physical parameters are calculated using the following logic, and the parameter calculation model is as follows: Here, the frequency resolution is clearly defined. The calculation method is as follows: ； in, Sampling rate, The number of FFT points; ； in, Down-conversion local oscillator frequency, Signal boundary index; ； in, Filter cutoff bandwidth : Protection bandwidth factor; Protection bandwidth factor The value is set to 1.2, determined based on filter transition band design principles. This aims to reserve a 20% bandwidth margin to accommodate the filter's roll-off characteristics and prevent amplitude attenuation at the passband edge. Based on this, resource allocation logic is executed; the module maintains a resource idle pool; when a new... When the signal disappears, i.e. the spectral entropy rises, the module requests an idle DDC channel ID from the pool and sends the above parameters to the channel via the control bus; if the signal disappears, i.e. the spectral entropy rises, the module reclaims the channel ID.

2. The shortwave broadband radio signal receiving and channelization processing system according to claim 1, characterized in that, The reconfigurable parallel channelization processing engine includes multiple parallel processing channels, each channel containing: A digital downconverter, the local oscillator frequency of which is dynamically configured by the dynamic channelization resource management module; A variable bandwidth filter bank, whose filter coefficients and decimation factors are generated in real time by the dynamic channelization resource management module based on the allocated bandwidth parameters.

3. The shortwave broadband radio signal receiving and channelization processing system according to claim 1, characterized in that, The spectrum analysis and signal detection module works continuously. When a drift in the center frequency of the tracked signal is detected, the dynamic channelization resource management module updates the local oscillator frequency parameters of the corresponding channel in real time to achieve tracking compensation for the signal frequency drift.

4. The shortwave broadband radio signal receiving and channelization processing system according to claim 1, characterized in that, The spectrum analysis and signal detection module and the dynamic channelization resource management module are integrated into a field-programmable gate array (FPGA), which utilizes its internal storage and arithmetic units to achieve high-speed parallel processing.

5. The shortwave broadband radio signal receiving and channelization processing system according to claim 1, characterized in that, Also includes: The spectrum occupancy assessment module is used to calculate the sparsity of signal activity within a wideband spectrum. The mode switching controller is used to enable the dynamic channelization resource management module when the spectrum is sparse, and switch to a fixed full-band scanning reception mode when the spectrum is dense.

Images