Energy efficiency adaptive regulation device and method for spectrum exploration equipment

Abstract

The application discloses an energy efficiency adaptive regulation device and method for a spectrum detection device, and relates to the technical field of radio spectrum monitoring. The device comprises a dynamic sleep-wake module, an energy efficiency adaptive regulation module, a task priority management module and a feedback optimization module. When on standby, low-rate sampling of environmental radio frequency signals is performed, and intensity, bandwidth and periodicity characteristics are extracted. Based on target characteristic determination, noise and missed detection are dynamically adjusted to trigger wake-up. When working, gain, sampling rate, quantization accuracy and algorithm complexity are jointly adjusted according to task requirements and energy efficiency constraints, and performance trade-off instructions are generated according to task priority when resources are limited. When idle, detection, power consumption and performance data are collected, and wake-up threshold and operating parameters are updated for the next period through reinforcement learning or regression analysis. Through the device and method, invalid wake-up can be inhibited, the risk of missed detection can be reduced, and adaptive balance of energy efficiency and task performance can be achieved.

Description

Technical Field

[0001] This invention relates to the field of radio spectrum monitoring technology, and in particular to an energy efficiency adaptive control device and method for spectrum detection equipment. Background Technology

[0002] As drones equipped with spectrum detection devices are increasingly used in electromagnetic environment monitoring, interference source localization, and communication support, these devices typically need to switch between long-term standby, intermittent operation, and emergency mission response. In these applications, spectrum detection devices must be sufficiently sensitive to changes in environmental radio frequency signals to avoid missing short-term or weak signal targets, while also being constrained by onboard power capacity and payload power consumption budgets, necessitating minimizing energy consumption during standby and operation. Therefore, achieving low-power operation while ensuring the quality of detection mission completion has become a long-standing key challenge in the engineering deployment of spectrum detection devices.

[0003] In existing technologies, low-power control of spectrum detection equipment often employs fixed-period wake-up or single-intensity threshold triggering for standby-working switching. This means the equipment initiates higher-power sampling and processing links according to a preset period, or is only triggered to wake up when the received signal strength reaches a fixed threshold. Once in working mode, the equipment typically operates with fixed configurations or minor rule adjustments to RF front-end gain, sampling rate, quantization accuracy, and algorithm complexity, lacking fine-grained matching between task requirements and operational status. Furthermore, when multiple tasks are concurrent or their urgency changes, existing solutions often use static priorities or simple preemptive scheduling, making it difficult to link task priorities with energy consumption control strategies, leading to a disconnect between resource allocation and energy efficiency regulation.

[0004] On the one hand, fixed-period wake-up and fixed-threshold triggering can lead to invalid wake-up or missed detection in complex electromagnetic environments, causing frequent start-stop or missing key target signals, thereby increasing energy consumption and reducing detection reliability. On the other hand, the lack of adaptive parameter adjustment based on task requirements and power constraints during operation means that when power is insufficient or multiple tasks are running in parallel, the system may either be overly conservative, causing the task to fail to be completed on time, or maintain a high-performance configuration, resulting in a rapid decline in battery life. This can affect the continuity and safety boundaries of UAV missions, and may even prevent key spectrum detection tasks from being prioritized under resource-constrained conditions. Summary of the Invention

[0005] To address the challenge of achieving low-power wake-up and energy-efficiency adaptive closed-loop control of spectrum detection equipment's operating parameters while ensuring the reliable completion of high-priority spectrum detection tasks under complex electromagnetic environments and resource-constrained conditions, this invention proposes an energy-efficiency adaptive control device and method for spectrum detection equipment.

[0006] The present invention achieves the above objectives through the following technical solutions:

[0007] An energy efficiency adaptive control device for a spectrum detection equipment, comprising:

[0008] The dynamic sleep / wake-up module is used to perform low-rate sampling and envelope detection of ambient radio frequency signals when the spectrum detection device is in standby mode, in order to obtain signal parameters characterizing signal strength, spectral bandwidth, and periodic characteristics. Based on the signal parameters, it performs target signal feature determination and dynamically adjusts the wake-up decision threshold based on the ambient noise level and historical missed detections. When the signal strength in the signal parameters reaches the dynamically adjusted wake-up decision threshold and the target signal feature determination is successful, a wake-up command is generated to activate the spectrum detection device to switch from standby mode to working mode.

[0009] An energy efficiency adaptive adjustment module is used to acquire the current task requirements when the spectrum detection device is in operation, and automatically adjust the operating control parameters of the spectrum detection device according to the current task requirements, including the RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy, and signal processing algorithm complexity.

[0010] The task priority management module is used to store pre-set task priority rules and monitor the operating status of the spectrum detection device. When the operating status meets the preset resource constraints, it generates a performance trade-off instruction according to the task priority rules and outputs it to the energy efficiency adaptive adjustment module to ensure that high-priority spectrum detection tasks are completed first.

[0011] The feedback optimization module is used to collect signal detection data, power consumption data and task performance data during the operation of the spectrum detection device, analyze and process them during idle periods, calculate the optimized wake-up decision threshold and operation control parameters through reinforcement learning algorithm or regression analysis algorithm, and apply them to the next operating cycle to improve the adaptability of the energy efficiency control strategy of the energy efficiency adaptive control device.

[0012] As a preferred embodiment of the present invention, the dynamic sleep / wake-up module includes:

[0013] The signal preprocessing unit is used to perform low-rate sampling and envelope detection on the ambient radio frequency signal when the spectrum detection device is in standby mode, and extract the signal strength parameter, spectrum bandwidth parameter and periodic characteristic parameter that constitute the signal parameters.

[0014] The feature determination unit is used to perform nonlinear matching calculations on the spectral bandwidth parameter and periodic feature parameter in the signal parameters based on a pre-stored target signal feature template, to generate bandwidth matching degree parameters and periodic matching degree parameters. Based on the bandwidth matching degree parameters and periodic matching parameters, a comprehensive matching degree parameter is generated according to a preset weighted fusion rule. The unit then outputs a target signal feature determination result and a corresponding feature confidence state based on the positional relationship of the comprehensive matching degree parameter relative to at least two preset matching threshold intervals. The target signal feature determination result is used to characterize whether the target signal feature determination is valid, and the feature confidence state is used to characterize the degree of matching between the signal parameters and the target signal feature template.

[0015] The threshold dynamic adjustment unit is used to update and adjust the wake-up decision threshold within a preset time window based on the target signal feature determination result, feature confidence state, and corresponding time distribution.

[0016] As a preferred embodiment of the present invention, the bandwidth matching parameter is calculated according to the following formula: ;

[0017] The periodic matching degree parameter is calculated according to the following formula: ;

[0018] In the formula, For bandwidth matching parameters, For periodic matching degree parameters; , These are the spectral bandwidth parameters and periodic characteristic parameters extracted by the signal preprocessing unit, respectively. , These represent the reference bandwidth parameter and reference period parameter stored in the target signal feature template, respectively. This represents the minimum allowed bandwidth threshold for the target signal. This indicates the maximum allowed period threshold for the target signal; , Indicates the matching attenuation coefficient; , This represents the penalty coefficient for exceeding the boundary. , It is a smoothing factor; Represents an exponential function;

[0019] The overall matching degree parameter is calculated according to the following formula: ;

[0020] In the formula, This is a comprehensive matching parameter; This is the consistency penalty coefficient; Indicates the bandwidth fusion weight index Modulated bandwidth matching contribution; Indicates the periodic fusion weight index Modulated period matching contribution;

[0021] When the comprehensive matching degree parameter falls into the first matching threshold range, the target signal feature determination result is valid, and the feature confidence state is marked as the first preset level;

[0022] When the comprehensive matching degree parameter falls into the second matching threshold range, the target signal feature determination result is invalid, or the feature confidence state is marked as the second preset level.

[0023] In a preferred embodiment of the present invention, the wake-up decision threshold satisfies the following mapping update relationship: ;

[0024] ;

[0025] In the formula, Indicates the first The wake-up decision threshold corresponding to each update cycle; This indicates the updated wake-up decision threshold; Indicates the first Threshold adjustment amount for each update cycle; This represents a truncation function used to truncate... Limit to the lower limit of the wake-up decision threshold Upper limit of wake-up decision threshold between; Indicates the first The comprehensive matching degree parameter corresponding to each update cycle; This indicates the threshold for a high match. This indicates a low matching threshold, and satisfies... ; Indicates the basic adjustment factor in the first direction; Indicates the basic adjustment factor for the second direction; Indicates the time distribution gating factor; This indicates an indicator function that takes the value 1 if the condition within the parentheses is true, and 0 otherwise.

[0026] When the target signal feature determination is successful, and the feature confidence state is at the first preset level, and the time distribution gating factor... When representing a concentrated distribution, the wake-up decision threshold is adjusted in the first direction;

[0027] When the target signal feature determination is not valid, or the feature confidence state is at the second preset level, and the time distribution gating factor... When characterizing a discrete distribution, the wake-up decision threshold is adjusted in a second direction opposite to the first direction.

[0028] As a preferred embodiment of the present invention, the dynamic sleep-wake module further includes an independently configured ultra-low power wake-up receiver for listening to wake-up calls from external control signals when the spectrum detection device is in standby mode.

[0029] When the ultra-low power wake-up receiver receives an external wake-up signal that matches the device address or device identifier of the spectrum detection device, the ultra-low power wake-up receiver generates an external wake-up trigger signal to directly trigger the spectrum detection device to switch from standby state to working state.

[0030] The signal preprocessing unit and the ultra-low power wake-up receiver work in parallel in standby mode, enabling the spectrum detection device to enter wake-up mode through an environmental radio frequency signal trigger path or an external control signal trigger path.

[0031] As a preferred embodiment of the present invention, the energy efficiency adaptive adjustment module includes:

[0032] The task requirement parsing unit is used to parse the current task requirements and generate a set of task requirement parameters that includes at least detection accuracy requirements, response latency requirements, and continuous working duration requirements.

[0033] The energy efficiency constraint assessment unit is used to generate a corresponding energy efficiency constraint parameter set based on the current remaining power status of the spectrum detection device, historical power consumption statistics, and the task requirement parameter set, which is used to limit the selectable value range and power consumption budget boundary of the operation control parameters in the current task cycle.

[0034] The parameter joint decision unit is used to perform joint selection and configuration of the operation control parameters based on the task requirement parameter set and the operation status characterization parameter set representing the current operation status of the spectrum detection device, provided that the energy efficiency constraint parameter set is satisfied; wherein, the operation status characterization parameter set includes at least the current channel environment index, algorithm processing load status and historical wake-up frequency parameters.

[0035] As a preferred embodiment of the present invention, the joint selection and configuration in the parameter joint decision-making unit includes:

[0036] While satisfying the requirements for detection accuracy and response delay, the energy efficiency cost function is minimized under the constraints of the set of operating state characterization parameters, and the variation of the operating control parameters within adjacent task cycles is limited to a preset stability constraint threshold.

[0037] The operation control parameter configuration results that match the current task requirements are generated by discrete parameter combination search or strategy mapping, and the operation control parameter configuration results are sent to the corresponding RF front-end, analog-to-digital converter and signal processing unit;

[0038] The energy efficiency cost function is defined as follows: ;

[0039] In the formula, This represents the energy efficiency cost function value used for joint parameter decision-making; This represents the vector of runtime control parameters for the current task cycle. This represents the vector of runtime control parameters corresponding to the previous task cycle; Indicated in the operation control parameter vector The power consumption cost item below; This represents the risk and cost of detection. Indicates the delay cost; This represents the stationarity cost term; The weighting parameters are respectively the power consumption cost weight, the detection risk cost weight, the latency cost weight, and the stability cost weight.

[0040] As a preferred embodiment of the present invention, the task priority management module includes:

[0041] The priority quantization unit is used to map the spectrum detection task to be executed to the corresponding task priority parameter based on the pre-defined task priority rules.

[0042] The operation status assessment unit is used to acquire and assess the operation status of the spectrum detection device in real time. The operation status includes at least the remaining power status, the number of current parallel tasks, and the change in the urgency of each spectrum detection task, and generates a corresponding set of operation status assessment parameters.

[0043] The performance trade-off decision unit is used to make a joint decision on the resource allocation relationship between different spectrum detection tasks based on the task priority parameter and the set of operating status evaluation parameters when the set of operating status evaluation parameters meets the preset resource constraints, and generate a performance trade-off instruction.

[0044] The performance trade-off instruction is used to determine the performance concession range corresponding to different spectrum detection tasks according to the size of the task priority parameter, and instruct the energy efficiency adaptive adjustment module to perform performance degradation, resource compression or delay operation on the operation control parameters corresponding to low priority spectrum detection tasks, while maintaining or improving the operation control parameter configuration corresponding to high priority spectrum detection tasks.

[0045] Furthermore, the performance trade-off instruction is at least used to adjust the weight parameters in the energy efficiency cost function, or to limit the selectable value range of the operation control parameters, so that the energy efficiency adaptive adjustment module reflects the task priority differences in the parameter joint decision-making process.

[0046] As a preferred embodiment of the present invention, the feedback optimization module includes:

[0047] The data collection unit is used to collect signal detection data, power consumption data and task performance data according to the task cycle or preset time window during the operation of the spectrum detection device, and to perform time alignment and feature processing on the signal detection data, power consumption data and task performance data to form a historical operation dataset for strategy evaluation.

[0048] The strategy evaluation unit is used to evaluate the energy efficiency performance of the dynamic sleep-wake module, the energy efficiency adaptive adjustment module, and the task priority management module under the current strategy configuration based on the historical operation dataset when the spectrum detection device is idle, using reinforcement learning algorithms or regression analysis algorithms, to generate corresponding strategy evaluation results and calculate the credibility index of the strategy evaluation results.

[0049] The parameter update unit is used to perform update calculations on the wake-up decision threshold and the operation control parameters when the credibility index meets the preset credibility conditions, and apply the updated wake-up decision threshold and the operation control parameters to the next operation cycle.

[0050] The parameter update unit is configured to introduce at least one update constraint mechanism during the update process. The update constraint mechanism includes historical parameter stability constraint, update step size constraint, or update cycle separation constraint, which is used to limit the change range of the wake-up decision threshold and the operation control parameters in adjacent operating cycles.

[0051] An energy efficiency adaptive control method for spectrum detection equipment includes:

[0052] When the spectrum detection device is in standby mode, it performs low-rate sampling and envelope detection on the ambient radio frequency signal to extract signal parameters characterizing signal strength, spectral bandwidth, and periodicity. Based on the signal parameters, it performs target signal feature determination and dynamically adjusts the wake-up decision threshold based on the ambient noise level and historical missed detections. When the signal strength in the signal parameters reaches the dynamically adjusted wake-up decision threshold and the target signal feature determination is successful, a wake-up command is generated to switch the spectrum detection device from standby mode to working mode.

[0053] When the spectrum detection device is in operation, the current task requirements are obtained, and the operating control parameters of the spectrum detection device are automatically adjusted according to the current task requirements, including the RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy, and signal processing algorithm complexity.

[0054] The system stores pre-defined task priority rules and monitors the operating status of the spectrum detection device. When the operating status meets preset resource constraints, it generates a performance trade-off instruction according to the task priority rules to ensure that high-priority spectrum detection tasks are completed first.

[0055] During the operation of the spectrum detection device, signal detection data, power consumption data, and task performance data are collected and analyzed during idle periods. The optimized wake-up decision threshold and operation control parameters are calculated using reinforcement learning algorithms or regression analysis algorithms and applied to the next operating cycle.

[0056] The beneficial effects of this invention are: it can suppress invalid wake-ups caused by non-target signals while maintaining the ability to respond to target signals, and reduce the risk of missed detection in complex electromagnetic environments through threshold adaptation, thereby mitigating the problems of energy waste and decreased detection reliability caused by fixed-cycle and fixed-threshold triggering. In the working state, an energy efficiency adaptive adjustment module is introduced to acquire and analyze the current task requirements, and automatically adjust operating control parameters such as RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy, and signal processing algorithm complexity. Simultaneously, the task priority management module generates performance trade-off instructions based on task priority rules when the preset resource constraints are met, and applies them to the energy efficiency adaptive adjustment process, ensuring that high-priority spectrum detection tasks are prioritized in situations of insufficient power or multiple tasks running in parallel. Furthermore, the feedback optimization module collects signal detection, power consumption, and task performance data during idle periods, and iteratively calculates the optimized wake-up decision threshold and operating control parameters using reinforcement learning or regression analysis algorithms, applying them to the next cycle. Through the synergistic mechanism of standby wake-up adaptation, operating parameter adaptation, priority-driven trade-off, and feedback iterative optimization, a dynamic balance between detection performance and energy consumption can be achieved under resource-constrained conditions, improving the adaptability of long-term energy efficiency control strategies and the stability of task completion. Attached Figure Description

[0057] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0058] Figure 1This is a schematic diagram of the modular structure of the device of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the described embodiments of the present invention are within the scope of protection of the present invention.

[0060] like Figure 1 As shown, this is an embodiment of the present invention. This embodiment provides an energy efficiency adaptive control device for a spectrum detection device, which is integrated inside the spectrum detection device and includes at least a dynamic sleep-wake-up module, an energy efficiency adaptive adjustment module, a task priority management module, and a feedback optimization module.

[0061] In this embodiment, the spectrum detection device includes at least the following two operating states:

[0062] Standby mode: The main RF front-end, high-speed ADC, and high-complexity signal processing algorithms are in the off or lowest power mode, with only the low-power sampling link or wake-up receiving link required by the dynamic sleep-wake module remaining.

[0063] Operating status: The main system of the equipment is awakened, and the RF front-end, analog-to-digital converter, and signal processing module are put into operation according to the current task requirements.

[0064] The state switching is uniformly triggered by the dynamic sleep / wake-up module and executed by the main system's power management unit or system control logic. When the spectrum detection device is in standby mode, the dynamic sleep / wake-up module performs low-rate sampling and envelope detection of the ambient radio frequency signal to obtain signal parameters characterizing signal strength, spectral bandwidth, and periodic characteristics. Based on the signal parameters, it performs target signal characteristic determination and dynamically adjusts the wake-up decision threshold based on the ambient noise level and historical missed detections.

[0065] The dynamic sleep / wake-up module includes at least a signal preprocessing unit, a feature determination unit, and a threshold dynamic adjustment unit. These sub-units can be deployed in the same low-power processing unit, or they can be implemented collaboratively by hardware circuitry and embedded software.

[0066] The signal preprocessing unit only activates the low-power sampling link in standby mode. Typical implementations include:

[0067] Low-speed ADC (e.g., 0.5–2MHz);

[0068] Envelope detection or simplified power estimation circuit;

[0069] Simple statistics module (mean, peak value, energy estimation).

[0070] The following signal parameters are output periodically or in an event-driven manner:

[0071] Signal strength parameter: used for subsequent wake-up decision;

[0072] Spectral bandwidth parameter and periodic characteristic parameter: used as inputs for determining the characteristics of the target signal.

[0073] In standby mode, the low-power sampling link acquires intermediate frequency / baseband envelope or low-resolution IQ data at a low sampling rate (1MHz in the example). Spectrum bandwidth parameters. If high-precision spectrum analysis is not required, the following low-complexity effective bandwidth estimation procedure can be used:

[0074] Within a preset frequency band, perform short-time energy spectrum estimation on the sampled data (using segmented FFT or multi-point Goertzel).

[0075] Calculate the noise reference The noise floor is obtained by statistically analyzing the mean or median of the power spectrum during periods without signal.

[0076] Setting the spectral decision threshold ,in Typical value is 3–10 dB (can be updated by the feedback optimization module);

[0077] The power spectrum higher than The continuous frequency range is mapped to the "occupied frequency band", and the width of the occupied frequency band is... (Unit: kHz)

[0078] This method has low computational complexity, is friendly to low-power MCUs, and yields... It is sufficient to support template matching.

[0079] Periodic characteristic parameters Used to describe the repetition period of a target signal pulse / burst. Implementation can employ an envelope triggering + timestamp differential process:

[0080] Set a trigger threshold for the envelope detection output. (The threshold can be the noise mean plus a certain number of standard deviations).

[0081] When the envelope exceeds When a "pulse event" is formed, record the event timestamp sequence. ;

[0082] Within a preset time window (e.g., 0.5–10 seconds, which can be consistent with the preset time window), calculate the interval between adjacent events. ;

[0083] Several Perform robust statistics (median or truncated mean) to obtain periodic characteristic parameters. .

[0084] The above calculations do not rely on complex correlation operations and are suitable for low-power implementation; when the target signal exhibits stable periodicity, It converges relatively quickly.

[0085] After receiving the signal parameters, the feature determination unit performs nonlinear matching calculations on the spectral bandwidth parameter and the periodic feature parameter based on the pre-stored target signal feature template to generate bandwidth matching degree parameters and periodic matching degree parameters. Based on the bandwidth matching degree parameters and the periodic matching degree parameters, a comprehensive matching degree parameter is generated according to a preset weighted fusion rule. Then, based on the positional relationship of the comprehensive matching degree parameter relative to at least two preset matching threshold intervals, the unit outputs the target signal feature determination result and the corresponding feature confidence state. The target signal feature determination result characterizes whether the target signal feature determination is valid, and the feature confidence state characterizes the degree of matching between the signal parameters and the target signal feature template. The output result is simultaneously provided to the threshold dynamic adjustment unit and the subsequent wake-up decision logic.

[0086] The target signal feature template must contain at least the following parameters and be stored in non-volatile memory (Flash / EEPROM) or a configuration file:

[0087] Reference bandwidth parameters Typical effective bandwidth of the target signal (e.g., 10–30 kHz);

[0088] Reference period parameters Typical period of the target signal (e.g., 20–50 ms);

[0089] Minimum bandwidth threshold Used to eliminate narrowband interference (e.g., 10kHz);

[0090] Maximum period threshold Used to eliminate long-period interference (e.g., 50ms or 200ms, depending on the application).

[0091] Calculate coefficients The initial value or allowable range (which can be calibrated during the manufacturing stage or updated by the feedback optimization module); where , This represents the matching attenuation coefficient, used to control the sensitivity of the matching degree to the decrease when the parameter deviates from the reference value; , This represents the out-of-bounds penalty coefficient, used to apply additional suppression to parameters that exceed the physical constraints of the target signal; , This is a smoothing factor used to avoid zero denominators and to adjust the slope of the matching function near the reference value; , These are the bandwidth fusion weight index and the period fusion weight index, respectively. This is the consistency penalty coefficient.

[0092] Template parameters can be obtained in two ways:

[0093] Offline configuration method: Pre-configured by engineers according to the business scenario (UAV spectrum monitoring task, target signal standard);

[0094] Online calibration method: During initial deployment, guide the device to collect known target signal samples and take... , The statistical mean / median is used as , .

[0095] Bandwidth matching parameter Calculate using the following formula: ;

[0096] The engineering implementation is recommended to be performed in two steps to avoid floating-point errors and ensure that the range is controllable:

[0097] (1) Similarity decay term

[0098] calculate , then calculate ; To avoid When the denominator is very small, it is unstable, which is acceptable in engineering. Or fix a small constant (e.g., 0.5kHz); The decay rate is determined by a typical initial value of 1–10; the larger the value, the more sensitive it is to deviations.

[0099] (2) Penalties for crossing the boundary

[0100] like Calculate the penalty amount and calculate ;like ,make ; Typically, it is taken as 0.5–1.5; when hour, much smaller It will significantly reduce the matching degree.

[0101] final During implementation, it is possible to... Truncate to [0,1].

[0102] Periodic matching degree parameter Calculate using the following formula: ;

[0103] Calculation method and Similarly, the difference lies in the direction of the boundary crossing: when Only then will the punishment be triggered. Desirable Or a fixed small constant (e.g., 1ms); Initial values ​​are suggested to be 1–10. Initial values ​​are recommended to be 0.5–1.5.

[0104] The above structure ensures a high degree of matching for cases close to the reference value through the exponential term, while further suppressing cases that violate physical / business constraints through the penalty term.

[0105] Comprehensive matching parameters Calculate using the following formula: ;

[0106] In the formula, Represents an exponential function;

[0107] The project implementation is recommended in three steps:

[0108] (1) Contribution modulation

[0109] Calculate the bandwidth matching contribution after exponential weight modulation. Contribution matching period Index weight Used to adjust the dominance of bandwidth and cycle time during fusion. Typical settings:

[0110] If the business relies more on bandwidth characteristics, then take (like );

[0111] If it relies more on periodic characteristics, then take ;

[0112] (2) Consistency penalty

[0113] Calculate the difference Calculate the consistency term ; The larger the value, the stronger the suppression of inconsistencies where one dimension matches well and the other dimension matches poorly; typical initial values ​​can be 1–10.

[0114] (3) Overall output

[0115] calculate And can Cut off to (0,1).

[0116] This fusion approach has clear feasibility: exponentiation and power operations can be implemented using lookup tables, fixed-point approximation, or MCU math libraries; for ultra-low power MCUs, piecewise approximation functions can be used, and the specific approximation method does not need to be specified in the manual.

[0117] Will Compare with at least two preset matching threshold intervals, for example:

[0118] First matching threshold range: ,in A value of 0.7–0.9 is acceptable.

[0119] Second matching threshold range: ,in A value of 0.2–0.5 can be taken, and .

[0120] When implemented,

[0121] when The output target signal feature determination result is "established", and the feature confidence state is marked as the first preset level;

[0122] when Output the target signal feature determination result as "not valid" or mark the feature confidence state as the second preset level;

[0123] when It can be used as an intermediate state for flexible processing of subsequent threshold update strategies.

[0124] The output should include at least: target signal feature determination results, feature confidence status, and comprehensive matching degree parameters. This is used by the threshold dynamic adjustment unit.

[0125] The threshold dynamic adjustment unit operates on a fixed update cycle, for example, triggering an update every 100ms or 1s. Each update uses cached data within a preset time window.

[0126] Cache comprehensive matching sequence ;

[0127] Cache the target signal feature determination result sequence (true / false);

[0128] Timestamp sequence of cached events .

[0129] The cache can be implemented using a circular queue, with a window length of 1–10 seconds.

[0130] The following are time distribution gating factors Calculation method:

[0131] Retrieve all established event timestamps within the window. Calculate the event interval ;

[0132] Calculate the mean with standard deviation (Or use the mean absolute deviation (MAD));

[0133] Calculate the dispersion index ,in Avoid dividing by zero for small constants;

[0134] Mapping the dispersion to a gating factor, for example: ,in This is a calibration constant (which can be 1 or set according to sample statistics).

[0135] The more concentrated the events, the smaller the interval fluctuations. The smaller, The closer to 1, the more discrete the events. The larger, The closer it is to 0.

[0136] It should be noted that the above are only preferred methods. It is only necessary to reflect the property of central convergence to 1 and discrete convergence to 0; the specific mapping function can be selected by the engineering implementation.

[0137] The threshold dynamic adjustment unit operates according to the following steps in each update cycle:

[0138] Read the current wake-up decision threshold Read the current Read the gate factor ;

[0139] Determine the conditions for a high degree of matching Low matching degree condition ; This indicates the high matching threshold. This indicates a low matching threshold, and satisfies... ; acceptable , ;

[0140] Calculate threshold adjustment amount :

[0141] ;

[0142] Update the threshold and truncate: ;

[0143] In the formula, Indicates the first The wake-up decision threshold corresponding to each update cycle; This indicates the updated wake-up decision threshold; Indicates the first Threshold adjustment amount for each update cycle; This represents a truncation function used to truncate... Limit to the lower limit of the wake-up decision threshold (e.g., -90dBm) and the upper limit of the wake-up decision threshold (e.g., between -30dBm); Indicates the first The comprehensive matching degree parameter corresponding to each update cycle; Indicates the basic adjustment factor in the first direction; This represents the basic adjustment coefficient in the second direction, if the threshold is... Expressed in dBm, it is recommended , Take 0.5–5 (corresponding to an adjustment of 0.5–5dB per cycle, which is actually smaller after gating); This indicates an indicator function that takes the value 1 if the condition within the parentheses is true, and 0 otherwise.

[0144] When the target signal feature determination is successful, and the feature confidence state is at the first preset level, and the time distribution gating factor is... When representing a concentrated distribution, the wake-up decision threshold is adjusted in the first direction;

[0145] When the target signal feature determination is invalid, or the feature confidence state is at the second preset level, and the time distribution gating factor is not valid, When characterizing a discrete distribution, the wake-up decision threshold is adjusted in a second direction, opposite to the first direction.

[0146] Comprehensive matching parameters The change indirectly reflects the impact of environmental noise level on the target signal feature matching result, and the time distribution gating factor. The changes in the value of reflect the statistical characteristics of historical missed detections in the time distribution dimension, by... and By jointly introducing an update process for the wake-up decision threshold, this embodiment can achieve comprehensive perception and response to environmental noise levels and historical missed detections without adding an additional high-power monitoring module.

[0147] The first and second direction adjustments are used to define the numerical update direction of the wake-up decision threshold. The first direction adjustment is defined as follows: when the high matching degree condition is met and the time distribution characterizes the target signal features as concentrated, the wake-up decision threshold is changed along a preset first numerical direction. The second direction adjustment is defined as follows: when the low matching degree condition is met or the feature confidence state is low, and the time distribution characterizes the target signal features as discrete, the wake-up decision threshold is changed along a second numerical direction opposite to the first numerical direction.

[0148] When the signal strength parameter reaches the dynamically adjusted wake-up decision threshold and the target signal characteristic determination is successful, a wake-up command is generated to activate the spectrum detection device to switch from standby to working state.

[0149] In one preferred embodiment, the dynamic sleep-wake module further includes a separately configured ultra-low power wake-up receiver for listening to wake-up calls from external control signals when the spectrum detection device is in standby mode.

[0150] The dynamic sleep / wake-up module contains at least two parallel links in standby mode:

[0151] Low-power sampling link (ambient RF signal triggering path): Antenna / RF front end → Low-power sampling link (low-speed ADC or envelope detection) → Signal preprocessing unit → Feature determination unit → Threshold dynamic adjustment unit → Wake-up command output;

[0152] Ultra-low power wake-up receiver link (external control signal trigger path): Antenna / coupler → Ultra-low power wake-up receiver → Address / identifier matching logic → External wake-up trigger signal output.

[0153] Both wake-up outputs are connected to the main system power management unit (PMU) or the main control wake-up pin (WakePin) to trigger the main system to power on.

[0154] The ultra-low power wake-up receiver can employ a simple modulation and demodulation structure (e.g., OOK / FSK envelope detection). The external wake-up signal frame structure includes at least: a preamble (synchronization), an address segment (device address / identifier), and a check segment (optional). The ultra-low power wake-up receiver continuously listens in standby mode. Upon detecting the preamble, it enters a short-time decoding state to decode the address segment bit sequence. The decoded address segment is then compared bit-by-bit with the pre-stored device address or device identifier; if they match, an external wake-up trigger signal is generated.

[0155] In standby mode, the signal preprocessing unit and the ultra-low-power wake-up receiver power on and operate simultaneously. The main system wake-up trigger uses OR logic: if any link generates a wake-up trigger signal, the main system enters the working state. When two links trigger simultaneously at close range, the main system only needs one wake-up; a wake-up latch can be set at the PMU level to avoid repeated power-ups. Furthermore, to prevent environmental links from continuing to consume resources after an external wake-up trigger, a standby link shutdown command can be issued by the main controller after the main system wakes up, shutting down the low-power sampling link or reducing the sampling frequency, and then restoring it after the task is completed.

[0156] The energy efficiency adaptive adjustment module is used to obtain the current task requirements when the spectrum detection equipment is in operation, and automatically adjust the operating control parameters of the spectrum detection equipment according to the current task requirements, including the RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy, and signal processing algorithm complexity.

[0157] The energy efficiency adaptive adjustment module consists of a task requirement analysis unit, an energy efficiency constraint evaluation unit, and a parameter joint decision-making unit, and interacts with each other via an internal bus or shared storage area. When the dynamic sleep / wake-up module triggers the device to enter the working state, the energy efficiency adaptive adjustment module is activated and operates in the following sequence within each task cycle:

[0158] Analyze the requirements of the current spectrum detection mission;

[0159] Assess available energy and power consumption constraints;

[0160] Construct a set of parameters representing the running status;

[0161] Combined selection and configuration of execution parameters;

[0162] The parameter configuration results are sent to the corresponding hardware and algorithm modules;

[0163] Each task cycle corresponds to one parameter joint decision-making process.

[0164] The task requirement parsing unit is used to parse the current task requirements. These requirements can originate from a higher-level scheduling node, task queue, or system preset strategy. The task description must include at least: the target frequency band or spectrum range, the detection result output period, and the task execution duration. The task requirements are input into the task requirement parsing unit in structured data form and parsed into a set of task requirement parameters, which must include at least:

[0165] Detection accuracy requirement parameters: These represent the requirements of the spectrum detection task for detection probability, false detection rate, or minimum detectable signal strength, and can be expressed as scalar levels or numerical ranges.

[0166] Response latency requirement parameter: This represents the maximum allowable latency from the start of the task or the appearance of the signal to the delivery of the detection result, and can be expressed as a time parameter in milliseconds.

[0167] Continuous working duration requirement parameter: This parameter indicates the length of time that the spectrum detection equipment needs to work continuously within the current task cycle, and can be expressed in seconds or minutes.

[0168] The energy efficiency constraint assessment unit obtains the following information at the beginning of each task cycle:

[0169] Current remaining battery status: such as battery percentage or estimated available energy;

[0170] Historical power consumption statistics include average power consumption and peak power consumption over several recent task cycles;

[0171] The set of task requirement parameters output by the task requirement parsing unit.

[0172] Based on the above inputs, the energy efficiency constraint assessment unit executes the following processing flow:

[0173] Calculate the allowable energy budget for the current task cycle based on the remaining battery status and the required continuous working time;

[0174] Based on historical power consumption statistics, estimate the power consumption range under different combinations of operating control parameters;

[0175] A set of energy efficiency constraint parameters is generated to limit the allowable power consumption budget boundary and the selectable range of operating control parameters (RF front-end gain, sampling rate, quantization accuracy, and algorithm complexity) within the current task cycle. For example, when the remaining power is low or the task duration is long, the set of energy efficiency constraint parameters will limit the upper limit of the sampling rate and algorithm complexity; when the remaining power is sufficient, higher performance configurations are allowed. The set of energy efficiency constraint parameters is output to the joint parameter decision unit in the form of parameter boundaries or a set of selectable levels.

[0176] Before making a joint selection, the parameter joint decision-making unit will also obtain a set of operating state characterization parameters, which includes at least:

[0177] Current channel environment indicators: such as channel occupancy, interference level, or noise estimate obtained from spectrum detection results;

[0178] Algorithm load status: such as current processing queue depth, amount of data that can be processed per unit time, or processor utilization;

[0179] Historical wake-up frequency parameters: such as the number of wake-ups or duty cycle within a preset time window, used to reflect the recent activity level of the system.

[0180] The above set of operational status characterization parameters is used to describe the operational status of the spectrum detection equipment during the current mission cycle and participate in subsequent joint decision-making.

[0181] The joint parameter decision unit receives the following inputs: a set of task requirement parameters, a set of energy efficiency constraint parameters, and a set of operational status characterization parameters. Its output is a set of operational control parameter configurations, including: RF front-end gain configuration, analog-to-digital converter sampling rate and quantization accuracy configuration, and signal processing algorithm complexity or algorithm path configuration.

[0182] In this embodiment, to reduce computational complexity and ensure real-time performance, each operation control parameter is discretized into a finite set of gears:

[0183] RF front-end gain level set ;

[0184] Analog-to-digital converter sampling rate range ;

[0185] Quantization precision level set ;

[0186] Set of signal processing algorithm complexity levels ;

[0187] Any combination of operating control parameters is expressed as .

[0188] Within each task cycle, the joint parameter decision-making unit performs the following steps:

[0189] Candidate set selection: Based on the set of energy efficiency constraint parameters, the above-mentioned range set is trimmed to form a restricted candidate set;

[0190] Performance and latency prediction: for each candidate parameter combination The system predicts its detection performance and processing latency. In this embodiment, the prediction is achieved through a lightweight regression model, with the input to the regression model being a vector of operating control parameters. The outputs are: predicted probe performance and predicted response latency. The regression model can use linear regression or small-scale multinomial regression. Its model parameters are trained using offline collected "parameter combination - performance / latency" samples, and the training process does not involve deep neural networks.

[0191] Feasibility screening: Only candidate parameter combinations that meet the requirements for detection accuracy and response delay are retained to form a feasible set;

[0192] Energy efficiency cost function evaluation: Calculate the energy efficiency cost function value for each candidate combination in the feasible set;

[0193] Optimal selection and distribution: Select the parameter combination with the minimum energy efficiency cost function value and distribute it to the corresponding module.

[0194] In this embodiment, the energy efficiency cost function is defined as: ;

[0195] In the formula, This represents the energy efficiency cost function value used for joint parameter decision-making; This represents the vector of runtime control parameters for the current task cycle. This represents the vector of runtime control parameters corresponding to the previous task cycle; This indicates the operation of the control parameter vector. The power consumption cost term is used to characterize the energy consumption level of the spectrum detection equipment during the current mission cycle. It is estimated through a parameter-power consumption mapping table or regression model under parameter combinations. The power consumption is then normalized. This represents the risk cost term for detection, used to characterize the risk in the operating control parameter vector. The degree to which the detection performance deviates from the required detection accuracy is calculated based on the deviation between the predicted detection performance and the required detection accuracy. When the predicted performance is lower than the required accuracy, the cost increases. This represents the time delay cost term, used to characterize the time delay cost in the operating control parameter vector. The following steps are taken: First, detect the degree to which the response latency deviates from the required response latency, and calculate based on the deviation between the predicted processing latency and the required response latency. These represent the power consumption cost weight, detection risk cost weight, latency cost weight, and stability cost weight, respectively, and are used to adjust the relative influence of each cost item in the energy efficiency cost function. The initial values ​​can be set according to the application scenario during the device deployment stage, and can be dynamically adjusted by the feedback optimization module based on the historical task execution effect.

[0196] The stability cost term characterizes the magnitude of change in the operational control parameters between the current task cycle and the previous task cycle. Defined as:

[0197] ;

[0198] In the formula, These represent the RF front-end gain configuration values ​​for the current task cycle and the previous task cycle, respectively. These represent the analog-to-digital converter sampling rate configuration values ​​for the current task cycle and the previous task cycle, respectively. These represent the quantization precision configuration values ​​for the current task cycle and the previous task cycle, respectively. These represent the complexity levels of the signal processing algorithms in the current task cycle and the previous task cycle, respectively. These represent the maximum and minimum allowable values ​​for the RF front-end gain, respectively. These are the maximum and minimum allowed sampling rates for the analog-to-digital converter, respectively. These are the maximum and minimum allowable values ​​for quantization precision, respectively. A distance metric representing the levels of algorithm complexity, used to characterize the differences between different algorithm complexity configurations; This represents the normalized upper limit of the distance between the algorithm complexity levels; This is the variation magnitude weighting coefficient, used to adjust the relative influence of different operating control parameters in the stationarity cost term;

[0199] when When the preset stationarity constraint threshold is exceeded, the stationarity cost term increases;

[0200] The joint decision-making unit can apply stability constraints in any of the following ways:

[0201] Hard constraint method: Directly eliminate candidate parameter combinations whose changes exceed a preset threshold;

[0202] Soft constraint method: Weighted penalty is applied to the energy efficiency cost function through a stationarity cost term.

[0203] The stability constraint threshold can be set during the equipment deployment phase, or it can be dynamically adjusted by the feedback optimization module based on operational data;

[0204] The final selected operational control parameter configuration results were simultaneously sent to:

[0205] RF front-end module, used to configure parameters such as gain;

[0206] The analog-to-digital converter module is used to configure the sampling rate and quantization accuracy;

[0207] The signal processing unit is used to select the corresponding algorithm path or algorithm complexity level.

[0208] The task priority management module stores pre-defined task priority rules and monitors the operating status of the spectrum detection equipment. When the operating status meets preset resource constraints, it generates performance trade-off instructions according to the task priority rules and outputs them to the energy efficiency adaptive adjustment module to ensure that high-priority spectrum detection tasks are completed first. It includes a priority quantization unit, an operating status evaluation unit, and a performance trade-off decision unit. These units can be implemented by different functional modules on the same processor or by independent control logic, and interact with each other through a shared memory area or an internal communication interface.

[0209] The priority quantization unit pre-stores task priority rules, including but not limited to:

[0210] The basic priority corresponding to the task type;

[0211] Priority weights corresponding to task sources;

[0212] Priority increment rules corresponding to task timeliness.

[0213] Task priority rules can be stored as lookup tables, rule tables, or parameterized functions. When there are spectrum detection tasks to be executed or currently being executed, the priority quantization unit maps each spectrum detection task to a corresponding task priority parameter according to the task priority rules. The task priority parameter can be a discrete level value or a continuous value, used to characterize the relative execution priority of different spectrum detection tasks under resource-constrained conditions. For example, high-urgency tasks can correspond to larger task priority parameters, while routine periodic monitoring tasks can correspond to smaller task priority parameters. The generated task priority parameters are output to the performance trade-off decision unit.

[0214] The operational status assessment unit is used to sense the operational status of the spectrum detection equipment in real time and generate a set of operational status assessment parameters.

[0215] The operational status assessment unit acquires at least three types of information:

[0216] Remaining battery status: such as current battery percentage or estimated available energy;

[0217] Current number of parallel tasks: This indicates the number of spectrum probe tasks that are currently executing or waiting to be executed.

[0218] Changes in task urgency: for example, a reduction in the remaining allowable time for the task, or an external instruction to increase the task's urgency.

[0219] The above information is normalized to generate a set of operational status evaluation parameters, which characterize whether the current state is resource-constrained. When the set of operational status evaluation parameters meets preset resource-constrained conditions (e.g., remaining battery power is below a threshold or the number of parallel tasks exceeds a preset upper limit), the performance trade-off decision unit is triggered to execute a performance trade-off decision.

[0220] After resource-constrained conditions are triggered, the performance trade-off decision unit executes the following processing flow:

[0221] Determining the performance concession range: Based on the set of task priority parameters and operational status evaluation parameters, the performance concession range corresponding to different spectrum detection tasks is determined; among them, high task priority parameters correspond to a smaller or zero performance concession range, low task priority parameters correspond to a larger performance concession range, and medium task priority parameters correspond to an intermediate level of performance concession range; the performance concession range is used to characterize the degree of performance degradation allowed for different tasks.

[0222] Generation of performance trade-off instructions: Generate performance trade-off instructions based on a determined performance concession amount, including at least one of the following or a combination thereof:

[0223] Instruct low-priority spectrum detection missions to perform performance degradation operations;

[0224] Instruct low-priority spectrum detection tasks to perform resource compression operations;

[0225] Instruct low-priority spectrum detection tasks to perform delayed or intermittent operations.

[0226] The performance trade-off command is output to the energy efficiency adaptive adjustment module for at least one of the following modes of operation:

[0227] Adjust the weighting parameters in the energy efficiency cost function so that low-priority tasks correspond to higher power consumption or performance costs in the joint parameter decision-making process;

[0228] Limit the range of selectable values ​​for the operation control parameters corresponding to low-priority tasks, such as limiting the upper limit of their sampling rate or algorithm complexity.

[0229] Through the above methods, the task priority management module and the energy efficiency adaptive adjustment module form a collaborative closed loop, enabling high-priority spectrum detection tasks to be completed first under resource-constrained conditions.

[0230] The feedback optimization module is used to collect signal detection data, power consumption data and task performance data during the operation of the spectrum detection equipment, analyze and process them during idle periods, calculate the optimized wake-up decision threshold and operation control parameters through reinforcement learning algorithm or regression analysis algorithm, and apply them to the next operating cycle to improve the adaptability of the energy efficiency control strategy of the energy efficiency adaptive control device.

[0231] The feedback optimization module is located in the control system of the spectrum detection equipment and includes a data acquisition unit, a strategy evaluation unit, and a parameter update unit. The feedback optimization module maintains data connections with the dynamic sleep / wake-up module, the energy efficiency adaptive adjustment module, and the task priority management module, but does not directly control the RF front-end or signal processing unit. Its effect is reflected by updating the wake-up decision threshold and operating control parameters. The feedback optimization module continuously runs the data acquisition function during the normal execution of the spectrum detection task by the equipment, and triggers the strategy evaluation and parameter update process when the equipment is idle.

[0232] During the operation of the spectrum detection equipment, the data collection unit collects the following three types of data according to the task cycle or preset time window:

[0233] Signal detection data includes the number of detected signals, detection success rate, number of missed detections, and statistical results of comprehensive matching parameters, provided by the dynamic sleep-wake module and signal processing unit.

[0234] Power consumption data: including average power consumption, peak power consumption, or energy consumption estimates for each task cycle, provided by the power management unit or power consumption monitoring module;

[0235] Task performance data includes task completion latency, task completion rate, and whether a task is delayed or downgraded, and is provided by the task scheduling and execution module.

[0236] Because the data sources and sampling frequencies differ, the data aggregation unit first performs time alignment processing on the data, mapping it to a unified task cycle or time window. Subsequently, it performs feature processing on the aligned data, for example:

[0237] Calculate statistical characteristics (mean, variance, extreme values) for continuous power consumption data;

[0238] Calculate performance indicators (detection rate, false negative rate) from the detection data;

[0239] Calculate the completion efficiency index based on the task performance data.

[0240] The processed data is organized into a historical running dataset and stored in memory for use by the policy evaluation unit.

[0241] The strategy evaluation unit operates only during periods of idle or low load for the spectrum detection equipment to avoid impacting real-time spectrum detection tasks. It uses historical operational datasets output by the data aggregation unit as input to evaluate the energy efficiency performance of the dynamic sleep / wake-up module, energy efficiency adaptive adjustment module, and task priority management module under the current strategy configuration.

[0242] In this embodiment, the policy evaluation unit can implement policy evaluation in any of the following ways:

[0243] (1) Evaluation method based on regression analysis: linear regression or multinomial regression model is adopted; the model input is the characteristic data in the historical running dataset, including power consumption characteristics, detection performance characteristics and task performance characteristics, and the model output is the comprehensive energy efficiency evaluation value corresponding to the current strategy configuration; the regression model is trained offline with historical sample data in the early stage of equipment deployment or maintenance stage, the model parameters are stored in the equipment, and only forward calculation is performed during runtime.

[0244] (2) Evaluation methods based on reinforcement learning

[0245] State definition: A feature vector of running states composed of historical running datasets;

[0246] Action definition: Fine-tune the current policy configuration or leave it unchanged;

[0247] Reward definition: Composed of both the reduction in power consumption and the improvement in task performance;

[0248] Learning method: Reinforcement learning algorithms are only used for policy evaluation and do not directly control hardware actions.

[0249] Regardless of the algorithm used, the strategy evaluation unit will calculate a reliability index for the strategy evaluation results based on at least one of the following factors to determine whether the current evaluation results are reliable:

[0250] Has the number of data samples reached the preset threshold?

[0251] The degree of consistency of evaluation results across different task cycles;

[0252] The stability of the improvement in energy efficiency performance before and after the strategy adjustment.

[0253] The parameter update unit performs parameter updates only if the following conditions are met simultaneously:

[0254] The credibility index output by the strategy evaluation unit meets the preset credibility conditions;

[0255] The equipment is about to enter its next operating cycle.

[0256] The parameter update unit only updates the wake-up decision threshold and operation control parameters (such as RF front-end gain, sampling rate, quantization accuracy, and algorithm complexity). The updated parameters are written to the corresponding module and take effect in the next running cycle.

[0257] To ensure system stability, the parameter update unit introduces at least one update constraint mechanism during the update process:

[0258] (1) Historical parameter stability constraint: The difference between the updated parameter and the parameter of the previous operating cycle is limited to a preset range to prevent drastic fluctuations in parameters.

[0259] (2) Update step size constraint: Set the maximum step size for the parameter adjustment range of each update to ensure that the parameters gradually converge.

[0260] (3) Update cycle separation constraint: Set different update cycles for different types of parameters. For example, the wake-up decision threshold is allowed to be updated more frequently, while the running control parameters adopt a slower update cycle.

[0261] The above constraint mechanism ensures the stability and convergence of the wake-up decision threshold and operation control parameters during the adaptive optimization process.

[0262] Another embodiment of the present invention provides an energy efficiency adaptive control method for a spectrum detection device, comprising the following steps:

[0263] When the spectrum detection equipment is in standby mode, it performs low-rate sampling and envelope detection on the ambient radio frequency signal to extract signal parameters that characterize signal strength, spectral bandwidth, and periodicity. Based on the signal parameters, it performs target signal feature determination and dynamically adjusts the wake-up decision threshold based on the ambient noise level and historical missed detections. When the signal strength in the signal parameters reaches the dynamically adjusted wake-up decision threshold and the target signal feature determination is successful, it generates a wake-up command to switch the spectrum detection equipment from standby mode to working mode.

[0264] When the spectrum detection equipment is in operation, the current task requirements are obtained, and the operating control parameters of the spectrum detection equipment are automatically adjusted according to the current task requirements, including the RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy and signal processing algorithm complexity.

[0265] Store pre-defined task priority rules and monitor the operating status of the spectrum detection equipment. When the operating status meets the preset resource constraints, generate performance trade-off instructions according to the task priority rules to ensure that high-priority spectrum detection tasks are completed first.

[0266] During the operation of the spectrum detection equipment, signal detection data, power consumption data, and task performance data are collected. During idle periods, these data are analyzed and processed. Optimized wake-up decision thresholds and operation control parameters are calculated using reinforcement learning algorithms or regression analysis algorithms and applied to the next operating cycle.

[0267] In summary, this invention achieves dynamic threshold wake-up by using low-rate sampling and target signal feature determination in standby mode. In working mode, it jointly adjusts the RF front-end gain, sampling rate, quantization accuracy, and algorithm complexity based on task requirements, power consumption, and operating status. Simultaneously, it generates performance trade-off instructions based on task priorities and iteratively updates the wake-up threshold and operating control parameters using reinforcement learning or regression analysis during idle periods. This suppresses invalid wake-ups and reduces the risk of missed detections in resource-constrained scenarios, thereby improving the adaptability of energy efficiency control strategies and the stability of task completion.

[0268] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. An energy efficiency adaptive control device for a spectrum detection equipment, characterized in that, include: The dynamic sleep / wake-up module is used to perform low-rate sampling and envelope detection of ambient radio frequency signals when the spectrum detection device is in standby mode, in order to obtain signal parameters characterizing signal strength, spectral bandwidth, and periodic characteristics. Based on the signal parameters, bandwidth matching parameters and periodic matching parameters are generated, and a comprehensive matching parameter is generated based on the bandwidth matching parameters and periodic matching parameters. According to the positional relationship of the comprehensive matching parameter relative to at least two preset matching threshold intervals, target signal feature determination is performed, and the target signal feature determination result and the corresponding feature confidence state are output. The wake-up decision threshold is dynamically adjusted based on the ambient noise level and historical missed detections. When the signal strength in the signal parameters reaches the dynamically adjusted wake-up decision threshold and the target signal feature determination is successful, a wake-up command is generated to activate the spectrum detection device to switch from standby mode to working mode. The bandwidth matching degree parameter is calculated according to the following formula: ; The periodic matching degree parameter is calculated according to the following formula: ; In the formula, For bandwidth matching parameters, For periodic matching degree parameters; , These are the spectral bandwidth parameters and periodic characteristic parameters extracted by the signal preprocessing unit, respectively. , These represent the reference bandwidth parameter and reference period parameter stored in the target signal feature template, respectively. This represents the minimum allowed bandwidth threshold for the target signal. This indicates the maximum allowed period threshold for the target signal; , Indicates the matching attenuation coefficient; , This represents the penalty coefficient for exceeding the boundary. , It is a smoothing factor; Represents an exponential function; The overall matching degree parameter is calculated according to the following formula: ; In the formula, This is a comprehensive matching parameter; This is the consistency penalty coefficient; Indicates the bandwidth fusion weight index Modulated bandwidth matching contribution; Indicates the periodic fusion weight index Modulated period matching contribution; When the comprehensive matching degree parameter falls into the first matching threshold range, the target signal feature determination result is valid, and the feature confidence state is marked as the first preset level; When the comprehensive matching degree parameter falls into the second matching threshold range, the target signal feature determination result is invalid, or the feature confidence state is marked as the second preset level; The wake-up decision threshold satisfies the following mapping update relationship: ; ； In the formula, Indicates the first The wake-up decision threshold corresponding to each update cycle; This indicates the updated wake-up decision threshold; Indicates the first Threshold adjustment amount for each update cycle; This represents a truncation function used to truncate... Limit to the lower limit of the wake-up decision threshold Upper limit of wake-up decision threshold between; Indicates the first The comprehensive matching degree parameter corresponding to each update cycle; This indicates the threshold for a high match. This indicates a low matching threshold, and satisfies... ; Indicates the basic adjustment factor in the first direction; Indicates the basic adjustment factor for the second direction; Indicates the time distribution gating factor; This indicates an indicator function that takes the value 1 if the condition within the parentheses is true, and 0 otherwise. When the target signal feature determination is successful, and the feature confidence state is at the first preset level, and the time distribution gating factor... When representing a concentrated distribution, the wake-up decision threshold is adjusted in the first direction; When the target signal feature determination is not valid, or the feature confidence state is at the second preset level, and the time distribution gating factor... When characterizing a discrete distribution, the wake-up decision threshold is adjusted in a second direction opposite to the first direction; An energy efficiency adaptive adjustment module is used to acquire the current task requirements when the spectrum detection device is in operation, and automatically adjust the operating control parameters of the spectrum detection device according to the current task requirements, including the RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy, and signal processing algorithm complexity. The task priority management module is used to store pre-set task priority rules and monitor the operating status of the spectrum detection device. When the operating status meets the preset resource constraints, it generates a performance trade-off instruction according to the task priority rules and outputs it to the energy efficiency adaptive adjustment module to ensure that high-priority spectrum detection tasks are completed first. The feedback optimization module is used to collect signal detection data, power consumption data and task performance data during the operation of the spectrum detection device, analyze and process them during idle periods, calculate the optimized wake-up decision threshold and operation control parameters through reinforcement learning algorithm or regression analysis algorithm, and apply them to the next operating cycle to improve the adaptability of the energy efficiency control strategy of the energy efficiency adaptive control device.

2. The energy efficiency adaptive control device for a spectrum detection equipment according to claim 1, characterized in that, The dynamic sleep / wake-up module includes: The signal preprocessing unit is used to perform low-rate sampling and envelope detection on the ambient radio frequency signal when the spectrum detection device is in standby mode, and extract the signal strength parameter, spectrum bandwidth parameter and periodic characteristic parameter that constitute the signal parameters. The feature determination unit is used to perform nonlinear matching calculations on the spectral bandwidth parameter and periodic feature parameter in the signal parameters based on a pre-stored target signal feature template, to generate bandwidth matching degree parameters and periodic matching degree parameters. Based on the bandwidth matching degree parameters and periodic matching parameters, a comprehensive matching degree parameter is generated according to a preset weighted fusion rule. The unit then outputs a target signal feature determination result and a corresponding feature confidence state based on the positional relationship of the comprehensive matching degree parameter relative to at least two preset matching threshold intervals. The target signal feature determination result is used to characterize whether the target signal feature determination is valid, and the feature confidence state is used to characterize the degree of matching between the signal parameters and the target signal feature template. The threshold dynamic adjustment unit is used to update and adjust the wake-up decision threshold within a preset time window based on the target signal feature determination result, feature confidence state, and corresponding time distribution.

3. The energy efficiency adaptive control device for a spectrum detection equipment according to claim 2, characterized in that, The dynamic sleep-wake module further includes an independently configured ultra-low power wake-up receiver, used to listen for wake-up calls from external control signals when the spectrum detection device is in standby mode; When the ultra-low power wake-up receiver receives an external wake-up signal that matches the device address or device identifier of the spectrum detection device, the ultra-low power wake-up receiver generates an external wake-up trigger signal to directly trigger the spectrum detection device to switch from standby state to working state. The signal preprocessing unit and the ultra-low power wake-up receiver work in parallel in standby mode, enabling the spectrum detection device to enter wake-up mode through an environmental radio frequency signal trigger path or an external control signal trigger path.

4. The energy efficiency adaptive control device for a spectrum detection equipment according to claim 1, characterized in that, The energy efficiency adaptive adjustment module includes: The task requirement parsing unit is used to parse the current task requirements and generate a set of task requirement parameters that includes at least detection accuracy requirements, response latency requirements, and continuous working duration requirements. The energy efficiency constraint assessment unit is used to generate a corresponding energy efficiency constraint parameter set based on the current remaining power status of the spectrum detection device, historical power consumption statistics, and the task requirement parameter set, which is used to limit the selectable value range and power consumption budget boundary of the operation control parameters in the current task cycle. The parameter joint decision unit is used to perform joint selection and configuration of the operation control parameters based on the task requirement parameter set and the operation status characterization parameter set representing the current operation status of the spectrum detection device, provided that the energy efficiency constraint parameter set is satisfied; wherein, the operation status characterization parameter set includes at least the current channel environment index, algorithm processing load status and historical wake-up frequency parameters.

5. The energy efficiency adaptive control device for a spectrum detection equipment according to claim 4, characterized in that, The joint selection and configuration in the parameter joint decision-making unit includes: While satisfying the requirements for detection accuracy and response delay, the energy efficiency cost function is minimized under the constraints of the set of operating state characterization parameters, and the variation of the operating control parameters within adjacent task cycles is limited to a preset stability constraint threshold. The operation control parameter configuration results that match the current task requirements are generated by discrete parameter combination search or strategy mapping, and the operation control parameter configuration results are sent to the corresponding RF front-end, analog-to-digital converter and signal processing unit; The energy efficiency cost function is defined as follows: ; In the formula, This represents the energy efficiency cost function value used for joint parameter decision-making; This represents the vector of runtime control parameters for the current task cycle. This represents the vector of runtime control parameters corresponding to the previous task cycle; Indicated in the operation control parameter vector The power consumption cost item below; This represents the risk and cost of detection. Indicates the delay cost; This represents the stationarity cost term; The weighting parameters are respectively the power consumption cost weight, the detection risk cost weight, the latency cost weight, and the stability cost weight.

6. The energy efficiency adaptive control device for a spectrum detection equipment according to claim 5, characterized in that, The task priority management module includes: The priority quantization unit is used to map the spectrum detection task to be executed to the corresponding task priority parameter based on the pre-defined task priority rules. The operation status assessment unit is used to acquire and assess the operation status of the spectrum detection device in real time. The operation status includes at least the remaining power status, the number of current parallel tasks, and the change in the urgency of each spectrum detection task, and generates a corresponding set of operation status assessment parameters. The performance trade-off decision unit is used to make a joint decision on the resource allocation relationship between different spectrum detection tasks based on the task priority parameter and the set of operating status evaluation parameters when the set of operating status evaluation parameters meets the preset resource constraints, and generate a performance trade-off instruction. The performance trade-off instruction is used to determine the performance concession range corresponding to different spectrum detection tasks according to the size of the task priority parameter, and instruct the energy efficiency adaptive adjustment module to perform performance degradation, resource compression or delay operation on the operation control parameters corresponding to low priority spectrum detection tasks, while maintaining or improving the operation control parameter configuration corresponding to high priority spectrum detection tasks. Furthermore, the performance trade-off instruction is at least used to adjust the weight parameters in the energy efficiency cost function, or to limit the selectable value range of the operation control parameters, so that the energy efficiency adaptive adjustment module reflects the task priority differences in the parameter joint decision-making process.

7. The energy efficiency adaptive control device for a spectrum detection equipment according to claim 6, characterized in that, The feedback optimization module includes: The data collection unit is used to collect signal detection data, power consumption data and task performance data according to the task cycle or preset time window during the operation of the spectrum detection device, and to perform time alignment and feature processing on the signal detection data, power consumption data and task performance data to form a historical operation dataset for strategy evaluation. The strategy evaluation unit is used to evaluate the energy efficiency performance of the dynamic sleep-wake module, the energy efficiency adaptive adjustment module, and the task priority management module under the current strategy configuration based on the historical operation dataset when the spectrum detection device is idle, using reinforcement learning algorithms or regression analysis algorithms, to generate corresponding strategy evaluation results and calculate the credibility index of the strategy evaluation results. The parameter update unit is used to perform update calculations on the wake-up decision threshold and the operation control parameters when the credibility index meets the preset credibility conditions, and apply the updated wake-up decision threshold and the operation control parameters to the next operation cycle. The parameter update unit is configured to introduce at least one update constraint mechanism during the update process. The update constraint mechanism includes historical parameter stability constraint, update step size constraint, or update cycle separation constraint, which is used to limit the change range of the wake-up decision threshold and the operation control parameters in adjacent operating cycles.

8. An energy efficiency adaptive control method for a spectrum detection device, based on the energy efficiency adaptive control device for a spectrum detection device according to any one of claims 1-7, characterized in that, include: When the spectrum detection device is in standby mode, it performs low-rate sampling and envelope detection on the ambient radio frequency signal to extract signal parameters characterizing signal strength, spectral bandwidth, and periodicity. Based on the signal parameters, it performs target signal feature determination and dynamically adjusts the wake-up decision threshold based on the ambient noise level and historical missed detections. When the signal strength in the signal parameters reaches the dynamically adjusted wake-up decision threshold and the target signal feature determination is successful, a wake-up command is generated to switch the spectrum detection device from standby mode to working mode. When the spectrum detection device is in operation, the current task requirements are obtained, and the operating control parameters of the spectrum detection device are automatically adjusted according to the current task requirements, including the RF front-end gain, analog-to-digital converter sampling rate, quantization accuracy, and signal processing algorithm complexity. The system stores pre-defined task priority rules and monitors the operating status of the spectrum detection device. When the operating status meets preset resource constraints, it generates a performance trade-off instruction according to the task priority rules to ensure that high-priority spectrum detection tasks are completed first. During the operation of the spectrum detection device, signal detection data, power consumption data, and task performance data are collected and analyzed during idle periods. The optimized wake-up decision threshold and operation control parameters are calculated using reinforcement learning algorithms or regression analysis algorithms and applied to the next operating cycle.

Images