Fixed unmanned aerial vehicle out-of-band dual-tone monitoring suppression transmission system

Abstract

The invention relates to a fixed unmanned aerial vehicle out-of-band dual-tone monitoring suppression transmission system. The fixed unmanned aerial vehicle out-of-band dual-tone monitoring suppression transmission system comprises a broadband spectrum monitoring receiving link, an out-of-band dual-tone transmission link, a receiving and transmitting self-interference isolation and leakage monitoring module, a closed-loop adaptive optimization control module, a monitoring result and suppression effect evaluation packaging module and a safety reporting link. The system determines a target link working frequency band according to the monitored target link working center frequency and occupied bandwidth, two paths of narrow-band continuous wave signals with independently adjustable frequency and power are transmitted outside the frequency band, and monitoring, constraint and evaluation of a target link suppression process are realized by combining digital pre-distortion, online calibration and leakage measurement. Spectrum parameters, intermodulation products, signal quality changes and station states are packaged and reported, and the method is suitable for unmanned aerial vehicle transmission monitoring, testing and controlled suppression scenes under law permission or controlled conditions.

Description

Technical Field

[0001] This invention belongs to the field of UAV wireless transmission monitoring, testing and spectrum regulation technology. Specifically, it relates to a fixed UAV out-of-band dual-tone monitoring and suppression transmission system, and more particularly to a wireless transmission system scheme deployed at a fixed site that uses out-of-band dual-tone signals to achieve controlled suppression of link reception performance in shared frequency bands and extended monitoring frequency bands by utilizing the nonlinear response of the target receiver front end, while maintaining the continuity of spectrum monitoring and the traceability of results at the site. Background Technology

[0002] For assessing the performance suppression and transmission impact of target UAV wireless links, existing technologies mainly employ the following methods. The first method involves directly transmitting a broadband signal within the target's operating frequency band, reducing the signal-to-noise ratio of the target receiver to degrade link performance. The main drawbacks of this method are: it occupies a significant amount of spectrum resources within the target frequency band, resulting in substantial collateral impact on legitimate services operating in the same band; it creates complex compliance boundaries; and when monitoring and transmission are deployed at the same site, the transmitted signal significantly compresses the dynamic range of the local receiver, making continuous monitoring difficult.

[0003] The second type involves transmitting a tracking narrowband signal at the target's operating frequency. Although the spectrum occupancy is improved compared to the aforementioned schemes, it still directly generates radiation within the target's operating frequency band. When the target uses frequency-hopping spread spectrum communication, tracking delay and frequency estimation accuracy become bottlenecks, and there is also the problem of self-interference with the monitoring and receiving link of this station.

[0004] The third type is protocol layer injection, which injects pseudo-frames or control information for specific communication protocols. This method relies on the degree of parsing of the target communication protocol, has limited applicability to unknown protocols or encrypted communication, and focuses more on protocol processing, which is not conducive to highlighting the structural design of the fixed wireless transmission system itself.

[0005] In summary, all the above solutions share a common shortcoming: the signal directly falls within the target operating frequency band. This makes it difficult to simultaneously address the following issues: First, the collateral impact on legitimate services in the same frequency band is difficult to quantify and control accurately; second, when transmitting and receiving are shared, the station's monitoring and reception are significantly affected by its own transmitted signals, making it difficult to perform monitoring and operational functions in parallel; third, spurious emissions and intermodulation generated by high-power transmissions within the band may spread to a wider frequency band, increasing compliance risks; and fourth, the evidence recording and reporting mechanisms for the operational process are inadequate, making it difficult to meet regulatory audit requirements.

[0006] From the perspective of the physical characteristics of the receiver's RF front-end, the front-end devices of a real RF receiving system—including low-noise amplifiers, mixers, and filters—may exhibit nonlinear responses. When two signals with frequencies f1 and f2 simultaneously enter a nonlinear device, a series of intermodulation products will be generated at the device's output. Typical third-order intermodulation products can be expressed as 2f1-f2 and 2f2-f1, while fifth-order intermodulation products can be expressed as 3f1-2f2 and 3f2-2f1. If f1 and f2 are located outside the target receiver's operating frequency band, but their third-order intermodulation product frequencies happen to fall within the operating frequency band, an equivalent in-band intermodulation component will be formed inside the target receiver, leading to a decrease in receiver sensitivity (desensitization effect) and consequently a deterioration in link quality. This phenomenon has a clear physical basis in the field of receiver testing, and its intensity is related to factors such as the target receiver's third-order input cutoff point (IIP3), two-tone arrival power, and frequency spacing.

[0007] The aforementioned physical principles suggest a novel design approach for wireless transmission at fixed sites: instead of directly radiating broadband signals within the target operating frequency band, two carefully configured narrowband signals are transmitted outside the band, utilizing the nonlinear response of the target receiver to generate in-band intermodulation components. This approach offers the following potential advantages: the transmitted signals do not directly fall into the target operating frequency band, making it easier to constrain the direct radio frequency impact on legitimate services within that band; the spectral occupancy of narrowband signals is significantly less than that of broadband direct transmission; and the site's monitoring and receiving link can maintain normal operation through transmit / receive isolation and spurious emission control, thereby enabling parallel operation of monitoring and transmission.

[0008] However, translating the above principles into a deployable, fixed, integrated monitoring and suppression transmission system still faces the following technical challenges. First, the power amplifier in the local transmit link is itself a nonlinear device. When two out-of-band signals pass through the same power amplifier, the amplifier itself will generate intermodulation spurious signals. If these spurious signals fall into the target frequency band, they will reduce the purity of the monitoring reception. Therefore, the problem of accurately suppressing the spurious signals on the transmit side must be solved. Second, under co-station conditions, transmit leakage will directly enter the monitoring and receiving link, requiring sufficient isolation at multiple levels in the spatial, radio frequency, and digital domains. Third, the optimal selection of dual-tone frequency pairing and power allocation is a multivariate optimization problem constrained by spurious signals, and it needs to be quickly recalculated when the target link frequency changes. Traditional exhaustive search methods are insufficient to meet real-time requirements. Fourth, monitoring results and suppression effects need to be recorded and reported in a unified, traceable, and tamper-proof manner to meet regulatory audit requirements.

[0009] Therefore, there is an urgent need for a complete wireless transmission system solution that can integrate broadband spectrum monitoring, out-of-band dual-tone transmission, leakage control, closed-loop optimization, and security reporting functions at fixed sites.

[0010] It is worth noting that existing patents and literature have proposed some solutions from the perspectives of full-band detection, broadband compressed sensing detection, integrated monitoring and control platforms, and adaptive interference cancellation. However, the above solutions either only focus on the monitoring side without involving transmission link control, or use in-band transmission methods that are difficult to implement in parallel with monitoring, or focus on intermodulation cancellation within the receiver rather than out-of-band two-tone transmission design. In the field of receiver testing, two-tone intermodulation testing is a common method for characterizing device nonlinearity, but there is still a lack of systematic structural solutions for transforming this testing principle into an out-of-band two-tone monitoring and suppression transmission system at a fixed site, and for addressing issues such as co-site spurious emission control, real-time frequency optimization, and compliance evidence retention in engineering. Summary of the Invention

[0011] The purpose of this invention is to provide a fixed UAV out-of-band dual-tone monitoring and suppression transmission system to solve the problems of large spectrum occupation, difficulty in controlling the impact on legitimate services, and difficulty in parallel monitoring and transmission in existing in-band direct transmission solutions, as well as the technical problems such as local spurious emission control, transmit / receive isolation, real-time optimization and compliance auditing faced when engineering and deploying the out-of-band dual-tone intermodulation principle.

[0012] To achieve the above objectives, the technical solution proposed in this invention includes: a broadband spectrum monitoring and receiving link, used for segmented detection and parameter extraction of the target frequency band, and determining the target link's operating center frequency, occupied bandwidth, and target link operating frequency band; an out-of-band dual-tone suppression transmission link, used to generate two independently adjustable narrowband continuous wave signals located outside the target link's operating frequency band, to induce intermodulation desensitization in the target receiver; a transmit / receive interference isolation and leakage monitoring module, used to ensure the normal operation of the monitoring receiver in the transmit state; a closed-loop adaptive optimization control module, used to optimize dual-tone parameters in real time under spurious constraints; a monitoring result and suppression effect evaluation and encapsulation module, used to uniformly encapsulate the entire process data into auditable evidence objects; and a secure reporting link, used to sign and correct the results before sending them up.

[0013] In one embodiment of the present invention, the broadband spectrum monitoring receiving link includes a broadband antenna or antenna array, a reconfigurable preselection filter, a low-noise amplifier, an ADC with direct RF sampling or down-conversion, and an FPGA / SoC. The FPGA / SoC employs a multiphase filter bank to achieve efficient channelized processing, with each sub-channel independently performing energy detection, cyclic stationary feature extraction, and modulation identification. Preferably, it covers shared frequency bands such as 2400–2483.5MHz, 5150–5350MHz, and 5725–5850MHz, and can support the licensed monitoring band of 1430–1444MHz and an extended monitoring interface of 5030–5091MHz through an extended RF front-end.

[0014] In another embodiment, the dual-channel digital signal source of the out-of-band dual-tone suppression transmission link independently generates two baseband narrowband continuous wave signals, which are then combined at the power combiner after digital-to-analog conversion and up-conversion. The digital predistortion module uses a memory polynomial model to inversely model the AM-AM and AM-PM characteristics of the power amplifier, and performs joint pre-compensation on the two signals in the baseband. The online self-calibration link continuously updates the predistortion coefficients by coupling a feedback signal from the power amplifier output and using an indirect learning architecture, ensuring that the spurious power of the power amplifier output in the target operating frequency band remains below the protection limit.

[0015] In another embodiment, the closed-loop adaptive optimization control module uses the dual-tone frequencies f1 and f2 and the powers P1 and P2 as four-dimensional optimization variables. Under steady-state target conditions, a Bayesian optimization algorithm is used to search for the optimal parameter combination within the constrained feasible region based on a Gaussian process surrogate model. In each iteration, the equivalent desensitization amount of the target link is evaluated and the surrogate model is updated. When the target link frequency changes rapidly, the closed-loop adaptive optimization control module switches to a fast response mode based on a pre-calculated lookup table: the system pre-calculates and caches the optimal dual-tone parameters for each possible target frequency; upon detecting a change, it directly looks up the table and switches, with a response delay lower than a preset threshold. After the target frequency stabilizes, Bayesian optimization is resumed for fine-tuning.

[0016] Compared with the prior art, the present invention has at least the following beneficial effects: First, by adopting an out-of-band dual-tone strategy instead of an in-band direct transmission method, the transmitted signal does not fall directly into the target operating frequency band, reducing the spectrum occupation from wideband to two narrowband signals, making it easier to constrain the direct radio frequency impact on legitimate services in the same frequency band.

[0017] Secondly, by using digital predistortion and online self-calibration, the intermodulation spurious signals generated by the local power amplifier are suppressed below the protection limit, ensuring that the station's monitoring and reception can still detect weak target signals normally under the suppression transmission working state.

[0018] Third, by using closed-loop adaptive optimization to find the optimal dual-tone parameters in real time under spurious constraints, the balance between desensitization effect and compliance spurious control is achieved.

[0019] Fourth, by encapsulating evidence objects, digitally signing, timestamp solidification, and forward error correction, the monitoring results and suppression operation process are made traceable and auditable, meeting regulatory compliance requirements.

[0020] Fifth, monitoring and suppression are integrated at the same fixed site, eliminating the need to deploy separate detection and suppression equipment, thus reducing the complexity of engineering deployment.

[0021] Furthermore, this invention is adaptable to both single-station independent working mode and multi-station collaborative working mode. In multi-station collaborative mode, the platform can jointly determine the target location and link parameters based on the monitoring results of multiple stations, and coordinate the dual-tone parameters of each station to achieve a spatial collaborative suppression effect.

[0022] Furthermore, this invention has an engineering basis. Whether out-of-band dual-tone (TOT) can produce the expected link performance suppression depends on factors such as the nonlinearity of the target receiver front-end, arrival power, propagation loss, antenna gain, and the on-site electromagnetic environment; therefore, its effective coverage distance and intensity should be determined through simulation or field measurement in conjunction with specific equipment and scenarios, and should not be absolutely limited. Attached Figure Description

[0023] Figure 1 This is a schematic diagram of the overall technical architecture of a fixed UAV out-of-band dual-tone monitoring and suppression transmission system proposed in this invention; Figure 2 This is a schematic diagram of the technical architecture of the generation principle and frequency pairing relationship of the out-of-band two-tone intermodulation products in this invention; Figure 3 This is a schematic diagram of the digital predistortion and online self-calibration link control principle in this invention; Figure 4 This is a simulation data analysis diagram of the desensitization amount of the target receiver and the spurious suppression effect of DPD under different dual-tone power conditions in this invention; Figure 5 This is a data analysis diagram of the convergence process and constraint satisfaction of the closed-loop Bayesian optimization in this invention; Figure 6 This is a data analysis chart comparing the desensitization effects under different target receiver IIP3 conditions in this invention. Detailed Implementation

[0024] The technical solution of the present invention will be further described below with reference to the accompanying drawings. It should be understood that the following embodiments are used to illustrate the principles, structure, and implementation path of the present invention, and are not intended to limit the scope of protection of the present invention. Without departing from the overall concept of the present invention, those skilled in the art can make equivalent substitutions or adjust engineering parameters for each module.

[0025] A fixed out-of-band dual-tone (TOT) monitoring and suppression transmission system for unmanned aerial vehicles (UAVs) is deployed at fixed monitoring stations or in a localized surveillance network composed of several fixed stations. Unlike pure monitoring systems, this system, in addition to receiving and monitoring, also suppresses the target UAV link based on the intermodulation desensitization principle through an out-of-band dual-tone transmission link after detecting the target UAV link. The system is preferably deployed at locations with stable power supply and network backhaul conditions, such as airport perimeters, park boundaries, the perimeter of important venues, or communication equipment rooms, and is ensured to operate continuously for a long period through temperature control, lightning protection, and timing devices. The antenna system of the station includes independent monitoring and receiving antennas and suppression transmission antennas, with sufficient spatial isolation between them.

[0026] In actual deployment, the site first determines the priority monitoring frequency bands based on the regulatory tasks. According to my country's current civil unmanned aerial vehicle radio management regulations, systems that achieve remote control, telemetry, and information transmission functions through direct communication can use 1430–1444MHz, 2400–2476MHz, and 5725–5829MHz. Fixed sites typically prioritize monitoring 2400–2476MHz and 5725–5829MHz, while also covering background service frequency bands such as 2400–2483.5MHz, 5150–5350MHz, and 5725–5850MHz to identify overlap between target links and wireless LAN, Bluetooth, and point-to-point transmission services. Table 1 shows the publicly disclosed regulatory frequency bands and monitoring strategies.

[0027] Table 1 Publicly disclosed regulatory frequency bands and monitoring strategies like Figure 1 As shown, the system as a whole consists of a broadband spectrum monitoring and receiving link 101, an out-of-band dual-tone suppression transmission link 102, a transmit / receive interference isolation and leakage monitoring module 103, a closed-loop adaptive optimization control module 104, a monitoring result and suppression effect evaluation encapsulation module 105, and a security reporting link 106. The broadband spectrum monitoring and receiving link 101 is responsible for the detection, identification, and parameter extraction of the UAV link; the out-of-band dual-tone suppression transmission link 102, after target confirmation, generates two out-of-band narrowband continuous wave signals located outside the target link's operating frequency band to implement link suppression; the transmit / receive interference isolation and leakage monitoring module 103 ensures the normal operation of the monitoring and receiving system during transmission; the closed-loop adaptive optimization control module 104 optimizes the dual-tone parameters in real time under spurious constraints; the monitoring result and suppression effect evaluation encapsulation module 105 uniformly encapsulates the data throughout the entire process; and the security reporting link 106 performs signing, error correction, and reporting. Data interaction between the links and modules is achieved through an internal bus and shared storage, and the entire system is coordinated and operated by the site's main control unit.

[0028] The broadband spectrum monitoring receiver link 101 includes a broadband antenna or antenna array, a reconfigurable preselection filter, a low-noise amplifier, a downconversion mixer stage if necessary, an analog-to-digital converter (ADC), and an FPGA / SoC. In a preferred embodiment, the reconfigurable preselection filter employs a switchable subband structure, selecting one of the candidate passbands (20MHz, 40MHz, or 80MHz) at each step, thereby completing segmented polling of the broadband band at a practically achievable sampling rate. For subbands with high-density background traffic, the system can appropriately shorten the polling interval; for subbands suspected of containing candidate UAV signals, the system extends the dwell time and enables refined scanning.

[0029] The FPGA / SoC internally implements a polyphase filter bank (PFB) or FFT-based multichannel channelization processing. In the coarse scan stage, a shorter time window and shorter FFT length are used to ensure coverage speed; in the fine scan stage, a longer time window, longer FFT length, or higher overlap rate are used on candidate frequency bands to improve frequency resolution and energy statistical stability. The system can also overlay short-time waterfall plots, spectral peak trajectories, and burst envelope analysis in the fine scan stage, ensuring that instantaneous energy anomalies caused solely by WLAN management frames, Bluetooth frequency hopping, or occasional broadband bursts are not directly identified as UAV links. In the candidate extraction process, the system first generates a set of candidate spectral peaks based on the coarse scan results, and then performs time-domain burst detection, cyclic stationary feature extraction, frequency stability estimation, and bandwidth fitting on the candidate spectral peaks. If a candidate signal simultaneously meets multiple conditions in duty cycle, frequency dwell time, cyclic stationary features, or modulation fingerprint, it enters the target confirmation process; if it only meets a single energy anomaly condition, a low-confidence label is added for subsequent window verification. For confirmed drone targets, the system extracts key parameters such as center frequency, occupied bandwidth, modulation type, and frequency hopping timing. These parameters will be directly used for subsequent out-of-band dual-tone frequency pairing calculations.

[0030] like Figure 2 As shown, the frequency pairing principle of out-of-band two-tone intermodulation products is as follows. Let the operating center frequency of the target UAV receiver be fc, the first out-of-band signal frequency be f1, and the second out-of-band signal frequency be f2, with both f1 and f2 located in the out-of-band region of the target receiver. Due to the nonlinear characteristics of the receiver's front-end devices, when signals f1 and f2 arrive at the receiver input simultaneously, third-order intermodulation products can be formed within the devices, with typical frequency components being 2f1-f2 and 2f2-f1.

[0031] When one of the aforementioned intermodulation components falls into the target operating frequency band, it may create an equivalent in-band effect within the target receiver.

[0032] If f1 and f2 are chosen such that one of the aforementioned third-order intermodulation products falls within the operating frequency band of the target receiver, this intermodulation product will form an equivalent in-band intermodulation component within the target receiver, reducing the receiver's effective sensitivity. The power of the two out-of-band signals arriving at the target receiver's input, the target receiver's IIP3 specification, and the target channel bandwidth will all affect the strength of this intermodulation component and the final link performance changes.

[0033] In the embodiments described in the specification, the relevant power relationships can be estimated based on the receiver's third-order intermodulation approximation model. The specific values ​​should be determined in conjunction with device parameters, link loss, and test conditions.

[0034] When the intermodulation component exceeds the target receiver's noise floor by a certain margin, the receiver's effective sensitivity will increase, and the signal-to-noise ratio of the target link will decrease. When the desensitization reaches a corresponding threshold, the target communication link will experience an increase in bit error rate or a significant decrease in communication quality. In an exemplary embodiment, if the target link operates at 2440MHz, f1 and f2 located on either side of its out-of-band can be selected so that the third-order intermodulation product formed by the two falls into the target frequency band near 2440MHz.

[0035] The specific selection of out-of-band dual-tone frequency pairing must meet the following constraints: Condition 1: Neither f1 nor f2 falls within the passband of the target receiver's in-band channel filter, ensuring that the target receiver's front-end filter does not directly introduce the dual-tone signal into the baseband. Condition 2: f1 and f2 do not fall within critical frequency bands protected by regulations, such as aviation navigation bands and emergency communication bands. Condition 3: 2f1-f2 or 2f2-f1 falls within the target receiver's operating frequency band fc±BW / 2, where BW is the target channel bandwidth. Condition 4: The frequency spacing between f1 and f2 should not be too large to ensure that the third-order intermodulation products dominate rather than being diluted by fifth-order or higher-order products. Condition 5: The selection of f1 and f2 should ensure that both are within the effective radiation gain coverage of the station's transmitting antenna.

[0036] The hardware components of the out-of-band dual-tone suppression transmission link 102 include: a dual-channel direct digital frequency synthesizer (DDS) or digital-to-analog converter (DAC) as a digital signal source, independently generating baseband narrowband continuous wave signals corresponding to f1 and f2 respectively; a digital predistortion module jointly performing nonlinear pre-compensation on the two signals; the two signals are each converted to radio frequency by an upconverter, combined by a power combiner, amplified by a final-stage power amplifier, and then radiated by a transmitting antenna. The transmitting antenna can be a broadband omnidirectional antenna or an adjustable directional antenna, the latter of which can concentrate radiated energy to improve the effective coverage distance when the target azimuth is known. In a preferred embodiment, the frequency resolution of the dual-channel DDS is better than 0.1Hz, and the phase noise is better than -110dBc / Hz@10kHz offset, ensuring that the generated dual-tone signal has high-purity narrowband characteristics and avoiding additional spurious emissions from its own phase noise in the target frequency band. The power combiner adopts a Wilkinson power combiner or hybrid coupler structure, with the combining loss controlled within 0.5dB, while maintaining an isolation of not less than 25dB between the two ports to reduce inter-channel interference.

[0037] like Figure 3 As shown, the working principle of the digital predistortion module and the online self-calibration link is as follows. The input-output relationship of the power amplifier can be characterized using a memory polynomial model, which consists of multiple-order nonlinear terms and finite memory terms.

[0038] The digital predistortion module constructs an inverse model based on the aforementioned model to pre-compensate the two-tone signal in the baseband domain. The memory polynomial model can be expressed as: Where x(n) is the power amplifier input signal, y(n) is the power amplifier output signal, K is the nonlinear order (usually 5 to 9), M is the memory depth (usually 3 to 7 sampling points), and a_{k,m} are the model coefficients. The digital predistortion module constructs the inverse model of this model and performs pre-compensation on the baseband signal before it enters the DAC, so that the output signal after power amplification is as close as possible to the ideal linear amplification.

[0039] The online self-calibration link samples a small portion of the signal from the power amplifier output via a directional coupler. After down-conversion and ADC conversion back to the digital domain, it is compared with the baseband reference signal. An indirect learning architecture is employed—a post-distorter model is first trained using the power amplifier output as input and the baseband reference as the desired output. Then, the coefficients of the post-distorter are copied to the pre-distorter. This process can be continuously run online using the NLMS or RLS algorithm, enabling the pre-distortion coefficients to track the characteristic drift of the power amplifier caused by temperature changes, aging, or power back-off. Under two-tone signal conditions, the core indicator of DPD is its ability to suppress intermodulation spurious signals at the 2f1-f2 and 2f2-f1 frequencies at the power amplifier output. The data shown in Table 2 are illustrative simulation results used to illustrate the relationship between power back-off and DPD spurious suppression, and do not constitute an absolute limitation on the specific implementation effect.

[0040] Table 2. Schematic simulation results of stray suppression for digital predistortion under different power back-off conditions. The transmit / receive interference isolation and leakage monitoring module 103 addresses the core technical bottleneck of co-station monitoring and suppression. For spatial isolation, a distance of at least 1.5 meters is maintained between the transmitting and receiving antennas, and an RF absorbing barrier is placed between them to improve spatial isolation, typically achieving 40 to 60 dB. For circuit isolation, a reconfigurable preselective filter at the front end of the monitoring and receiving link provides additional out-of-band suppression at the dual-tone transmission frequency. For digital domain isolation, the leakage reference acquisition branch extracts transmitted signal samples from the output of the transmitting link through a directional coupler, and after delay and amplitude alignment, sends them to the FPGA / SoC for adaptive leakage cancellation, further reducing residual leakage by 15 to 25 dB. The combined effect of these three levels of isolation allows the monitoring and receiving link to maintain detection sensitivity for weak target signals even in out-of-band dual-tone transmission operation.

[0041] The real-time spurious spectrum measurement unit, within each optimization update cycle, utilizes the spectrum analysis capabilities of the monitoring and receiving link to scan and quantify the spurious power generated by the locally transmitted signal within the target frequency band. The measurement results include the total spurious power within the target frequency band, the peak spurious frequency and its power, and the margin between the spurious power and the protection limit. This data is fed back in real-time to the closed-loop adaptive optimization control module as constraint input, ensuring that the system pursues maximum desensitization while not violating the spurious protection limit.

[0042] The closed-loop adaptive optimization control module 104 operates as follows: Step S1_opt, Initialization: Based on the detected target link center frequency fc, calculate an initial frequency pair (f1, f2) according to the third-order intermodulation fall-in condition, and set the initial power (P1, P2) to a medium level. Step S2_opt, Transmission and Measurement: Set the initial parameters to the dual-channel digital signal source and start out-of-band dual-tone transmission; simultaneously monitor the target link signal quality change in the receiving link, and the spurious spectrum measurement unit measures the spurious power of the local unit in the target frequency band. Step S3_opt, Evaluation: Calculate the desensitization evaluation quantity and spurious constraint satisfaction under the current parameter combination. Step S4_opt, Optimization Update: Input the current observation sample into the Bayesian optimization surrogate model, and select the next set of candidate parameters within the constraint feasible region through the expected improvement quantity acquisition function. Step S5_opt, Convergence Judgment: If the change in N consecutive iterations is less than the convergence threshold, it is considered to have converged to the current optimal value; if the constraint is not satisfied, it reverts to the previous parameter combination that satisfies the constraint and narrows the search range. Step S6_opt, Frequency Hopping Tracking: If a frequency hop of the target link is detected, the corresponding parameters are immediately read from the pre-calculated lookup table and a fast switch is completed.

[0043] In one exemplary implementation, the Bayesian optimization surrogate model employs a Matérn 5 / 2 kernel Gaussian process, and the constraints are incorporated into the acquisition function using a probabilistic constraint method: the probability of constraint satisfaction is considered simultaneously when calculating the expected improvement, and candidate points are selected only in regions where the constraint satisfaction probability exceeds 0.95. The pre-computation lookup table is generated as follows: during system deployment or rule update phases, a set of candidate frequency pairs is generated at 1MHz intervals for each possible target frequency fc. For each pair, spurious power is predicted using a power amplifier nonlinear model, and a subset of parameters satisfying the constraints is selected. The combination with the largest estimated desensitization is then stored in the lookup table. The typical size of the lookup table is hundreds to thousands of entries, with a lookup response time of less than 1 millisecond, meeting the real-time requirements of frequency hopping tracking.

[0044] To illustrate the feasibility of out-of-band dual-tone desensitization, simulation analysis can be performed under different IIP3, dual-tone arrival power, link signal-to-noise ratio margin, and receiver bandwidth conditions to statistically analyze the target receiver sensitivity degradation or related link quality indicators. The data in Table 3 are illustrative simulation results used to illustrate the changing trends of desensitization under different receiver nonlinear conditions and do not constitute an absolute limitation on specific equipment.

[0045] Table 3. Schematic simulation results of desensitization parameters under different target IIP3 and dual-tone arrival power conditions. To illustrate the convergence performance of closed-loop optimization under spurious constraints, the Bayesian optimization can be iterated multiple times under the conditions of setting protection limits, frequency search range, and power adjustment range, and the change trajectories of desensitization and spurious power can be statistically analyzed. The data shown in Table 4 are illustrative simulation results used to illustrate that the optimization process has the possibility of convergence when the constraints are met, and do not constitute an absolute limitation on the specific numerical values.

[0046] Table 4. Schematic simulation results of the closed-loop Bayesian optimization convergence process. After completing one round of suppression operations, the monitoring result and suppression effect evaluation encapsulation module 105 performs a comparative evaluation of the signal quality of the target link before and after suppression. Evaluation indicators include: change in the detected power of the target link signal (dB), i.e., the difference in the detected power of the target signal at this station before and after suppression; change in carrier-to-noise ratio (dB), reflecting the trend of change in the target link signal quality; and change in frequency hopping success rate (percentage), reflecting the response effect of the frequency hopping communication target after being subjected to intermodulation. The monitoring result and suppression effect evaluation encapsulation module 105 uniformly writes the following parameters into the evidence object: target center frequency fc, occupied bandwidth BW, dual-tone frequency pairing (f1, f2), dual-tone transmit power (P1, P2), theoretical intermodulation product frequency and power estimate, desensitization evaluation quantity, measured local spurious emissions, suppression start time and duration, DPD status and coefficient version, station number and station clock status. Each evidence object also includes a unique object identifier UUID, a content digest hash SHA-256, and an audit index.

[0047] The secure reporting link 106 is used for the trusted transmission of evidence objects. At the site, the evidence object first generates a content digest hash (using the SHA-256 algorithm), then the secure chip uses the site's private key to complete a digital signature (using ECDSA or SM2 algorithm). Subsequently, a trusted timestamp (from GPS / BeiDou timing or NTP synchronization source) and forward error correction redundancy (using RS encoding or LDPC encoding) are appended before transmission to the platform. Upon receiving the evidence, the platform verifies the signature validity, temporal order, and error correction results, and writes the audit log to the hash chain or the distributed tamper-proof evidence storage module. If a short-term interruption occurs in network transmission, the site maintains a controlled cache queue locally with a capacity of no less than 1000 evidence objects. Once the link is restored, the original signed evidence objects are retransmitted in timestamp order instead of being regenerated, to maintain evidence consistency and temporal integrity. Upon receiving the retransmitted object, the platform confirms its consistency with the initial transmission attempt by comparing the signature and digest hash, and records the number of retransmissions and the link status at the time of retransmission in the audit log, ensuring a complete closed loop of the evidence chain.

[0048] In a further embodiment, the system also includes a full lifecycle security audit and controlled display module. The security audit module records in real time actions such as site configuration changes, frequency band switching, target detection triggering, dual-tone parameter calculation, transmission start / stop, DPD coefficient updates, spurious constraint out-of-bounds alarms, optimization iteration logs, evidence signing, and remote queries. Each record carries a precise timestamp, the operating entity, and the execution result, and is linked to preceding and following events through a hash chain, allowing for tracing the context of any operational decision during subsequent review. The controlled display module uses server-side rendering and pixel-stream feedback to display the spectrum, dual-tone configuration status, and suppression effect evaluation results, ensuring that the original IQ data and signature keys are always retained on the controlled side of the site.

[0049] To improve monitoring accuracy and reduce false alarm rates, the system maintains a background service template library at the site level. The template library stores long-term statistical features according to a multi-dimensional index based on sub-frequency bands, time periods, workday types, and environmental labels. These features include noise quantiles, spectral peak density, common bandwidth distributions, burst duration histograms, and frequency dwell time distributions. When a candidate signal enters the confirmation process, the system calculates the similarity between its feature vector and the background template. If the candidate signal is highly consistent with the background template, its alarm priority is reduced; if the candidate signal consistently deviates from the background template across multiple windows, its review priority is increased. In this way, the system avoids misclassifying isolated, legitimate service energy peaks as UAV targets. In an optional implementation, the background template library also supports online incremental updates. After the system has run for a preset period (e.g., one week or one month), it automatically performs weighted fusion of the background statistical features within that period with existing templates, enabling the templates to adapt to long-term, slow changes in the electromagnetic environment.

[0050] In multi-station collaborative mode, the platform can compare the detection results of different stations for the same candidate target, including center frequency, timestamp, modulation fingerprint, and burst period, for joint confirmation. Once the target is confirmed, the platform can coordinate multiple stations to implement out-of-band dual-tone suppression from different directions. The dual-tone parameters of each station are uniformly optimized by the platform to avoid the generation of new unexpected intermodulation products between the dual tones transmitted by multiple stations. If a station cannot achieve the ideal desensitization effect due to limited transmission conditions (such as tight spurious constraints), the platform can increase the power allocation of other stations to achieve overall collaborative suppression. In multi-station collaborative mode, the platform also needs to ensure that the transmission timing of each station remains synchronized to avoid phase mismatch of intermodulation products at the target end due to timing deviations, which would weaken the desensitization effect. To this end, each station achieves a time synchronization accuracy better than 100 nanoseconds through GPS / BeiDou co-view or IEEE 1588 precision clock protocol.

[0051] For UAV targets employing frequency-hopping spread spectrum communication, this invention achieves rapid tracking and switching of dual-tone frequency pairings through a pre-calculated lookup table. After detecting the target's frequency-hopping pattern, the system can predict the next frequency hopping frequency using a frequency-hopping sequence prediction algorithm and prepare the corresponding dual-tone parameters in advance from the lookup table, making the dual-tone frequency switching nearly synchronized with the target's frequency hopping. In a more advanced implementation, the system can simultaneously cover multiple frequency positions within the target's frequency hopping set using multiple pairs of dual tones. However, in this case, it is necessary to additionally evaluate whether the intermodulation products between multiple pairs of dual tones fall into the protected frequency band, and add corresponding constraints to the closed-loop adaptive optimization control module.

[0052] The site self-calibration logic runs automatically during low-traffic periods or after configuration changes, with a typical self-calibration process taking approximately 5 to 15 minutes. First, the monitoring receiver link is calibrated for local oscillator deviation, gain flatness, and noise figure by injecting an internal calibration signal with known frequency and power to measure the end-to-end transfer function of the receiver link. Second, the out-of-band dual-tone transmission link is calibrated for power: calibration signals with known power are independently transmitted to each of the two channels, the actual output power at the transmit antenna port is measured, and the power calibration table is updated to ensure dual-tone power control accuracy is better than ±0.5 dB. Third, the DPD module coefficients are retrained: dual-tone calibration signals are transmitted in the calibration band, and the power amplifier output spurious signals are observed through the feedback link, iteratively updating the DPD coefficients until the spurious suppression meets the target; typical DPD training convergence requires 100 to 500 iterations. Finally, the transmit / receive isolation is measured and evaluated by measuring the residual power leaked from the transmitter to the receiver in each monitoring sub-band. All calibration results are recorded in the site parameter area and written into the site health metadata of the evidence object. If any indicator exceeds the allowable range, the system will lower the site health score, add a maintenance pending label to the reporting object, and send a maintenance alarm notification to the platform.

[0053] During the long-term operation of a site, regulatory rules and optimization parameters may be adjusted due to policy updates or changes in the electromagnetic environment. To address this, this invention designs a rule version chain mechanism: each rule set includes applicable frequency bands, monitoring priorities, dual-tone power limits, spurious emission protection limits, optimization iteration limits, and reporting strategies. When the platform issues new rules to a site, the site first performs integrity verification and compatibility checks in the shadow configuration area, and only switches to the new rules after confirmation. If spurious emission exceeds limits or desensitization effects abnormally deteriorate after switching, the platform can immediately trigger a rollback to the previous stable version. This mechanism facilitates controlled evolution in large-scale site networks. The system generates a version identifier and digest hash for each rule set and records the rule's activation time, deactivation time, and affected site scope, forming a complete rule change chain in the audit log, allowing for post-event traceability of the impact of any parameter change on monitoring and suppression effectiveness.

[0054] The modules described in this invention can be implemented collaboratively by FPGA, DSP, general-purpose processor, and dedicated RF devices, or integrated with a system-on-a-chip and programmable logic. The digital predistortion module's operations can be implemented in hardware logic within the FPGA to meet real-time requirements, while Bayesian optimization can be implemented in software on an ARM processor or a standalone embedded processor. For the site-side software, either locally embedded programs or containerized components uniformly distributed and version-controlled by the platform can be used. Regardless of the implementation method, as long as it can complete spectrum monitoring, out-of-band dual-tone signal generation and transmission, DPD spurious suppression, transmit / receive isolation and leakage monitoring, closed-loop optimization, evidence encapsulation, and secure reporting, it should fall within the protection scope of this invention. In engineering deployment, the overall system can be integrated into a standard 19-inch rack or outdoor protective enclosure, including an RF front-end unit, digital processing unit, power amplifier unit, power supply and heat dissipation unit, and network communication unit. The typical total power consumption is 200 to 500W, meeting the long-term unattended operation requirements of fixed sites.

[0055] For sites requiring further enhanced spatial selectivity, this invention can also optionally incorporate multi-antenna arrays and simplified direction-of-arrival (DOA) estimation logic. In this case, the monitoring and receiving link not only outputs spectral parameters but also provides the approximate incident direction range of candidate signals; the suppression transmission link can utilize phased array antennas to achieve beam pointing, concentrating the dual-tone radiation energy in the target direction, thereby increasing the effective coverage distance and reducing collateral impacts on legitimate services in non-target directions. It should be understood that multi-antenna arrays and DOA estimation are beneficial extensions to this invention, rather than essential conditions for achieving out-of-band DOA suppression. With DOA estimation capability, the closed-loop adaptive optimization control module can also incorporate DOA information into the optimization variables, aligning the main lobe of the DOA beam with the target direction and the null point with the legitimate service direction, further reducing collateral impacts on the surrounding area.

[0056] In actual operation, when a station detects a suspected UAV remote control signal in the 2400–2483.5MHz frequency band, the system first identifies candidate spectral peaks during a coarse scan, and then confirms their duty cycle, frequency stability, and modulation characteristics in a fine scan, comparing them with surrounding Wi-Fi signals. The Fi service exhibits differences. Once the target is identified and meets legally permitted or controlled testing conditions, the closed-loop adaptive optimization control module calculates dual-tone frequency pairing based on the target's center frequency and occupied bandwidth. For example, if the target operates at 2440MHz, a pair of f1 and f2 signals can be selected on either side of its out-of-band frequency, causing the third-order intermodulation products formed by these two signals to fall into the target frequency band. The system then initiates out-of-band dual-tone transmission while simultaneously monitoring the receiving link to continuously track changes in the target signal quality to evaluate the desensitization effect. If the target uses frequency hopping communication to switch between 2440MHz and 2460MHz, the system detects the transition and immediately reads the pre-optimized dual-tone parameters corresponding to 2460MHz from the lookup table and completes a rapid switch, with a switching delay less than a preset threshold. Throughout the process, the DPD module continuously ensures that local spurious emissions within the target frequency band do not exceed the protection limit, and the evidence encapsulation module records the complete parameters and effect evaluation data for each round of suppression operations in real time.

[0057] In the 5725-5850MHz image transmission scenario, candidate signals typically have a wider occupied bandwidth and a more stable carrier frequency structure. The pairing space for dual-tone frequencies is relatively ample at this point, but frequency selection should still be constrained by local frequency licensing, protected band avoidance, and local spurious emission protection limits. In an exemplary and non-limiting implementation, if the rule base determines that the relevant frequency points do not belong to protected or prohibited bands, f1 can be selected near 5900MHz and f2 near 6050MHz, so that the third-order intermodulation products 2f1-f2 fall into the target image transmission frequency band near 5750MHz. It should be noted that the 5GHz band includes Dynamic Frequency Selection (DFS) related bands. When selecting dual-tone frequencies, the system must avoid falling into DFS bands such as 5250-5350MHz and 5470-5725MHz to prevent triggering the DFS radar detection mechanism of legitimate devices in the same frequency band, which could have unintended consequences. This constraint has been incorporated into the set of constraints for closed-loop optimization. In 5GHz image transmission suppression scenarios, since the target image transmission signal usually adopts OFDM or OFDM-like modulation, its bandwidth can reach 10MHz to 40MHz. A single continuous wave dual-tone pair mainly forms narrowband intermodulation components. The actual link quality degradation also depends on the blocking characteristics of the target receiver front end, the out-of-band suppression degree of the channel filter, and the link margin. For wider target channels, the system can use time-division sweeping, multi-tone coordination, or controlled modulation extension to increase the coverage range, but the multi-tone intermodulation products and protected frequency band constraints must be evaluated simultaneously.

[0058] When conducting power budget analysis from an engineering feasibility perspective, the fixed site transmit power, antenna gain, propagation loss, target receiver antenna gain, target receiver front-end IIP3, and on-site obstruction and multipath conditions all collectively determine the out-of-band two-tone arrival power and intermodulation product intensity. Due to significant differences in various UAV platforms, receiver front-end structures, and site environments, the effective range and suppression strength of this invention should be determined comprehensively based on specific equipment parameters, link budget, simulation results, and on-site test data; a simple, uniform numerical limit is not advisable.

[0059] This invention does not preclude collaboration with other passive monitoring capabilities. For example, stations can spatiotemporally correlate the spectrum monitoring results output by this invention with photoelectric identification, radar detection, remote identification information, or navigation data to form a more complete situational map. In the core link of this invention, all spectrum evidence is independently encapsulated, independently signed, and independently audited, ensuring that even if other sensing links are missing, the spectrum-side results still have independent verification value. Furthermore, when the system is in pure monitoring mode (without dual-tone transmission), the function and performance of its monitoring link are completely consistent with those of a traditional fixed spectrum monitoring station, and there will be no additional interference or performance degradation due to the presence of a dual-tone transmission link. This design allows the system to flexibly switch between pure monitoring mode and monitoring plus suppression mode according to the regulatory strategy, improving the system's adaptability in different regulatory scenarios. When dealing with major events or temporary support tasks, the system can also temporarily adjust parameters such as monitoring priority, dual-tone power limit, and spurious protection limit according to platform instructions. After the task is completed, it automatically restores the normal configuration and retains all parameter changes and operation records during the task.

[0060] Regarding environmental adaptability, this invention considers the stringent requirements of long-term outdoor deployment at fixed sites. The RF front-end devices must maintain stable performance across a wide temperature range of -40℃ to +55℃. The system monitors the power amplifier junction temperature in real time using temperature sensors, automatically reducing transmission power to protect device lifespan when the junction temperature approaches the preset upper limit. In high-temperature seasons or at high-altitude sites, the system can activate thermal derating mode to maintain basic suppression capabilities at lower power rather than shutting down due to overheating. For lightning protection, the antenna feeder system employs two levels of lightning protection at the antenna port and cabinet entrance, consisting of a gas discharge tube and a transient suppression diode, respectively, ensuring the RF path is not damaged by lightning strikes. The power supply system is equipped with an uninterruptible power supply (UPS) and surge protection devices, enabling at least 4 hours of emergency operation at the site during mains power outages, and automatically performing system self-checks and parameter recovery upon power restoration. Regarding fault tolerance, the system employs a watchdog timer and heartbeat monitoring mechanism for the core processing module. When a software anomaly or task timeout is detected, a module-level reset is automatically triggered instead of a full-site restart, minimizing monitoring blind spots caused by software failures. The transmit link is also equipped with real-time VSWR monitoring. When the VSWR at the antenna port exceeds a preset threshold, the power amplifier output is immediately shut down to prevent damage to the power amplifier due to abnormal antenna feeder connection. In addition, the system firmware supports remote online upgrades. The platform can push firmware update packages to the site through an encrypted secure channel. After completing integrity verification, the site automatically performs the upgrade during low-service periods. If the upgrade fails, it automatically reverts to the previous stable version.

[0061] This invention, through the coordinated operation of modules such as broadband spectrum monitoring and reception, out-of-band dual-tone transmission, digital predistortion spurious emission control, transmit / receive self-interference isolation, closed-loop adaptive optimization, evidence encapsulation, and secure reporting, achieves controlled suppression of the target link's reception performance without directly transmitting broadband signals within the target's operating frequency band, while maintaining the continuity and traceability of the station's spectrum monitoring results. This system can meet the needs of fixed stations for UAV link monitoring, controlled suppression, compliance auditing, and multi-station collaboration in complex electromagnetic environments with shared frequency bands. Compared to in-band direct transmission solutions, this invention has significant technical advantages in terms of spectrum occupancy, collateral impact constraints, parallel monitoring and transmission capabilities, and evidence integrity.

Claims

1. A fixed unmanned aerial vehicle (UAV) external dual-tone monitoring and suppression transmission system, characterized in that, include: A broadband spectrum monitoring and receiving link is used to perform segmented gating, radio frequency reception, analog-to-digital sampling, and multi-resolution spectrum analysis on a preset UAV wireless transmission frequency band to detect and extract the spectrum parameters of the target UAV remote control link or image transmission link. The spectrum parameters include the target link's operating center frequency, occupied bandwidth, and the target link's operating frequency band determined by the operating center frequency and occupied bandwidth. The broadband spectrum monitoring and receiving link sequentially includes a broadband antenna or antenna array, a reconfigurable preselection filter, a low-noise amplifier, an analog-to-digital converter (ADC) for direct radio frequency sampling or down-conversion, and an FPGA / SoC for performing multiphase filtering channelization, time-frequency analysis, and candidate signal screening. An out-of-band dual-tone transmission link is used to determine the operating frequency band of the target link based on the spectrum parameters extracted by the broadband spectrum monitoring and receiving link after detecting the target UAV link, and to generate a first narrowband continuous wave signal and a second narrowband continuous wave signal located outside the operating frequency band of the target link. The frequency and power of the first narrowband continuous wave signal and the second narrowband continuous wave signal can be independently adjusted, and after power amplification, they are radiated through a transmitting antenna. The frequency pairing relationship of the first narrowband continuous wave signal and the second narrowband continuous wave signal is set so that one of the third-order intermodulation products generated by the nonlinear response of the front end of the target UAV receiver falls into the operating frequency band of the target link, so as to implement controlled suppression of the target link's receiving performance and provide subsequent evaluation. The out-of-band dual-tone transmission link includes a dual-channel digital signal source, a digital predistortion module, an upconverter, a power amplifier, and a transmitting antenna. The transmit / receive interference isolation and leakage monitoring module is used to perform spatial isolation and / or circuit isolation between the transmit and receive links under the condition of co-location of monitoring reception and out-of-band dual-tone transmission, and to measure in real time the leakage power and spurious power generated by the transmitted signal of the out-of-band dual-tone transmission link in the monitoring reception frequency band, so as to maintain the continuous operation of the monitoring reception link. The transmit / receive interference isolation and leakage monitoring module includes a transmit / receive antenna isolation device, a leakage reference acquisition branch, and a spurious spectrum real-time measurement unit. The closed-loop adaptive optimization control module is used to implement closed-loop control on the frequency pairing and power allocation of the two narrowband continuous wave signals based on the target link parameter change trend, target link reception performance change, and local spurious power output by the spurious spectrum real-time measurement unit, so as to ensure that the target link reception performance suppression evaluation meets the preset requirements and that the local in-band spurious power does not exceed the preset protection limit. The monitoring results and suppression effect evaluation encapsulation module is used to encapsulate the target link's spectral parameters, detected intermodulation product frequency points and power, receiver performance change evaluation quantities, signal quality change indicators before and after suppression, and site operation status into an auditable evidence object. The secure reporting link is used to send the evidence object to the regulatory platform after digitally signing, timestamping, and forward error correction encoding.

2. The fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, The reconfigurable preselective filter is a preselective filter with switchable passband or adjustable center frequency, used to divide the frequency band to be monitored into multiple sub-bands, including the 2400-2483.5MHz, 5150-5350MHz, and 5725-5850MHz shared background monitoring sub-bands, as well as the 2400-2476MHz and 5725-5829MHz sub-ranges highly related to the direct communication of UAVs. It can also support the 1430-1444MHz licensed monitoring band and the 5030-5091MHz extended monitoring interface through the extended RF front end. The FPGA / SoC uses a polyphase filter bank PFB or fast Fourier transform FFT to realize multi-channel channelization processing, and independently performs energy detection, cyclic stationary feature extraction, frequency hopping timing analysis and modulation fingerprint recognition on each sub-channel to distinguish UAV remote control signals, image transmission signals and background service signals.

3. The fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, The frequency pairing of the first narrowband continuous wave signal and the second narrowband continuous wave signal is determined according to the following principle: Let the operating center frequency of the target UAV receiver be fc, the frequency of the first signal be f1, and the frequency of the second signal be f2. Then f1 and f2 are both located outside the operating frequency band of the target link, and one of the third-order intermodulation products formed by f1 and f2 falls within the operating frequency band of the target link near fc, while retaining an allowable frequency offset margin Δf. The dual-channel digital signal source independently generates the narrowband baseband signals corresponding to f1 and f2, which are converted to radio frequency by their respective digital-to-analog converters and upconverters and then combined and output at the power combiner. When the target link experiences a frequency jump or drift, the closed-loop adaptive optimization control module recalculates the frequency pairing that satisfies the intermodulation fall-in condition and updates the parameters of the dual-channel digital signal source.

4. A fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, The digital predistortion module uses a memory polynomial model or a generalized memory polynomial model to inversely model the nonlinear characteristics of the power amplifier. It performs pre-compensation before the first narrowband continuous wave signal and the second narrowband continuous wave signal enter the power amplifier, so as to suppress the third-order and fifth-order intermodulation spurious power generated at the output of the power amplifier in the target link operating frequency band to below the preset protection limit. The out-of-band dual-tone transmission link also includes an online self-calibration link. The online self-calibration link is used to couple a feedback signal from the output of the power amplifier and compare the feedback signal with the baseband reference signal. It uses an indirect learning architecture to update the predistortion coefficient in real time, so that the digital predistortion module can track the nonlinear characteristic drift of the power amplifier caused by temperature changes, aging, or power backoff.

5. A fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, The transmit / receive antenna isolation device includes spatial isolation between the transmit and receive antennas with a minimum preset spacing and an RF absorbing structure located between the transmit and receive antennas; the leakage reference acquisition branch extracts the transmitted signal sample from the output of the transmit link through a directional coupler, and inputs the sample to the FPGA / SoC of the broadband spectrum monitoring and receiving link after bandpass filtering and delay alignment, so as to perform transmit leakage cancellation processing in the digital domain; the spurious spectrum real-time measurement unit quantifies and measures the power of local transmit leakage and intermodulation spurious in the target frequency band in the monitoring and receiving link in each optimization update cycle, and uses the measurement result as the constraint input of the closed-loop adaptive optimization control module.

6. A fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, The closed-loop adaptive optimization control module uses the dual-tone frequency pairings f1 and f2 and the dual-tone power allocations P1 and P2 as optimization variables, the estimated equivalent desensitization amount of the target UAV receiver as the optimization objective, and the constraints of the local spurious power not exceeding the preset protection limit and the dual-tone frequency not falling into the protected frequency band within the target link's operating frequency band. A Bayesian optimization algorithm is used to construct a surrogate model, and the optimal parameter combination is searched within the constrained feasible region using the expected improvement amount acquisition function. When the target link frequency changes rapidly, the closed-loop adaptive optimization control module switches to a fast response mode based on a pre-calculated lookup table, directly reading the corresponding pre-optimized frequency pairings and power parameters from the lookup table according to the detected new target frequency. The Bayesian optimization mode is then restored after the target link frequency stabilizes.

7. A fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, The monitoring results and suppression effect evaluation encapsulation module compares and evaluates the signal quality of the target link before and after suppression. The evaluation indicators include at least one of the following: change in the detected power of the target link signal, change in the carrier-to-noise ratio, and change in the frequency hopping success rate. The encapsulation module standardizes and encapsulates the target center frequency, occupied bandwidth, dual-tone frequency pairing, dual-tone transmit power, intermodulation product frequency and power estimate, desensitization assessment quantity, local spurious measured value, suppression duration, site number and site clock status according to the preset evidence object description specification; each evidence object also includes a unique object identifier, digest hash and audit index.

8. A fixed UAV out-of-band dual-tone monitoring and suppression transmission system according to claim 1, characterized in that, It also includes a full lifecycle security audit and controlled display module, which is used to record operation logs in real time for each stage from spectrum scanning, target detection, dual-tone parameter calculation, transmission start-up, spurious measurement, optimization iteration to signature reporting; the controlled display module uses server-side rendering and pixel stream feedback to display the spectrum, dual-tone configuration status and suppression effect evaluation results to the monitoring terminal, so that the original IQ data and signature private key are always kept on the controlled side of the site; and the audit logs are solidified through hash chain or distributed anti-tampering evidence storage technology.

Images