An examination room area AI intelligent signaling shield detector management and control system

Abstract

The application discloses an examination room area AI intelligent signaling shield detector detection management and control system, belongs to the technical field of shield detector detection management and control, and comprises an examination room information extraction module, a shield instrument state monitoring module, a shield instrument state analysis module, an examination room signal detection module, a detection signal analysis module, an examination room shielding analysis control module and a feedback early warning terminal. The examination room information extraction module is used for extracting environment configuration information and shield instrument equipment file information of each examination room in a target examination site. The shield instrument state monitoring module is used for receiving a control instruction, sending an opening or closing signal to a shield instrument of a designated examination room, and collecting operation and maintenance state information of each shield instrument in real time and automatically. The shield instrument state analysis module is used for performing hierarchical analysis and quantitative evaluation on the health condition of the shield instrument. On the basis of realizing shield detector detection management and control, the application can also actively detect signals and dynamically adjust the output frequency.

Description

Technical Field

[0001] This invention relates to the field of jamming detection and control technology, and more specifically, to a detection and control system for an AI intelligent signaling jammer used in examination areas. Background Technology

[0002] When a mobile phone is working, it connects to a base station via radio waves within a specific frequency range, transmitting data and voice at a certain baud rate and modulation scheme. To counter this communication principle, a mobile phone signal jammer scans from the low end of the forward channel to the high end at a certain speed during operation. This scanning speed can create garbled interference in the received message signal, preventing the phone from detecting normal data transmitted from the base station and thus preventing it from establishing a connection. The phone will exhibit symptoms such as searching for a network, no signal, or no service.

[0003] Existing systems largely rely on examination staff manually turning on the jammers before the exam and turning them off afterward. This method carries the risk of human negligence or operational errors, such as forgetting to turn them on or accidentally turning them off, directly leading to jamming failure. The entire process lacks automated control and status verification mechanisms, resulting in lax management. Traditional jammers are treated as black-box devices; the system only cares whether they are powered on, but cannot know whether their internal modules are working properly or whether the emitted interference signals meet the standards. Most existing jamming systems only have signal interference transmission functions, which are passive defenses. They cannot detect whether there are actually cheating signals attempting to communicate in the examination room, nor can they distinguish between mobile phone standby signals and ongoing cheating data transmission. Summary of the Invention

[0004] To address the problems existing in the prior art, the purpose of this invention is to provide a detection and control system for an AI intelligent signaling jammer in examination areas. In addition to detecting and controlling the jammer, this invention can also actively detect signals and dynamically adjust the output frequency.

[0005] To solve the above problems, the present invention adopts the following technical solution: A detection and control system for an AI-powered intelligent signaling jammer in an examination room includes: an examination room information extraction module, a jammer status monitoring module, a jammer status analysis module, an examination room signal detection module, a detection signal analysis module, an examination room jamming analysis and control module, and a feedback and early warning terminal. The examination room information extraction module is used to extract environmental configuration information and shielding device file information of each examination room in the target examination site. The shielding device status monitoring module is used to receive control commands and send on or off signals to the shielding devices in the designated examination room, and to collect the operation and maintenance status information of each shielding device in real time and automatically. The shielding device status analysis module is used to perform hierarchical analysis and quantitative evaluation of the health status of the shielding device; The examination room signal detection module is used to continuously detect the radio signal environment in the examination room. The detection signal parsing module is used to parse the raw data acquired by the examination room signal detection module; The examination room shielding analysis and control module is used to make intelligent decisions based on multi-source information fusion, dynamically analyze and calculate the target shielding power of each examination room, and send instructions to the shielding device to adjust the power. The feedback early warning terminal is used to receive decisions from the examination room shielding analysis and control module and execute early warnings, while simultaneously realizing closed-loop management and collaborative handling of early warning information. As a preferred embodiment of the present invention, the examination room information extraction module includes an examination room environment and configuration information extraction unit and a shielding device file information extraction unit; The examination room environment and configuration information extraction unit is used to collect the spatial volume of the examination room, the attenuation coefficient of the building material and the distribution information of the surrounding communication base stations. The spatial volume and the attenuation coefficient of the building material are used by the examination room shielding analysis and control module to calculate the theoretically suitable shielding power threshold. The shielding device file information extraction unit is used to extract the shielding device's own history and status information.

[0006] As a preferred embodiment of the present invention, the shielding device status monitoring module includes an instruction control and execution unit, a multi-parameter data acquisition unit, a data preprocessing and caching unit, and a preliminary status feedback and communication interface unit. The instruction control and execution unit is used to receive remote instructions from the examination room shielding analysis control module and send control signals to the designated examination room signal jammer to turn it on or off. The multi-parameter data acquisition unit is used to collect the operation and maintenance status information of the target examination room signal jammer in real time. The operation and maintenance status information includes the electromagnetic field strength reflecting the shielding effect, the signal parameters of each shielding frequency band, the number of signals, and the signal amplitude. The data preprocessing and caching unit is used to perform preliminary processing on the raw operation and maintenance status data acquired by the multi-parameter data acquisition unit. The preliminary processing includes data cleaning, format unification and timestamp marking, and is organized and temporarily stored according to the three-dimensional structure of examination room identification, monitoring time point and signal frequency band. The preliminary status feedback and communication interface unit is used to package the pre-processed and structured operation and maintenance status information and transmit it to the backend shielding device status analysis module in real time through the network interface. At the same time, the preliminary status feedback and communication interface unit is also used to receive query instructions from the shielding device status analysis module or the examination room shielding analysis control module and return the corresponding real-time or historical data.

[0007] As a preferred embodiment of the present invention, the shielding device status analysis module includes a usage health assessment unit, an operation and maintenance health assessment unit, and a comprehensive health analysis and decision-making unit; The health assessment unit is used to assess the historical wear and tear or aging of the shielding device due to long-term use based on its historical usage data. The operation and maintenance health assessment unit is used to assess the current working efficiency of the shielding device based on the real-time monitoring data after it is activated. The current working efficiency includes the electromagnetic field effect generated by the device and its ability to suppress target communication signals. The integrated health analysis and decision-making unit is connected to the usage health assessment unit and the operation and maintenance health assessment unit. The integrated health analysis and decision-making unit is used to integrate the historical wear and tear assessment results with the current work efficiency assessment results to generate a unified, quantifiable and comparable health assessment index. Based on the comparison result of the health assessment index and the preset threshold, it triggers the diagnostic decision and subsequent control process for the shielding device. The operation and maintenance health assessment unit includes a shielded magnetic field compliance calculation subunit and a shielded signal compliance calculation subunit; The shielding magnetic field compliance calculation subunit is used to calculate a compliance index reflecting the magnetic field generation capability of the shielding device based on real-time monitored electromagnetic field strength data. The shielding signal compliance calculation subunit is used to calculate the compliance index reflecting the suppression effect of the shielding device on the target signal based on the signal parameters of each target shielding frequency band monitored in real time. The shielding magnetic field compliance calculation subunit locates the lowest electromagnetic field strength and the statistically obtained magnetic field strength reduction rate from the real-time monitoring data, and compares them with the preset normal reference value. The operation and maintenance health assessment unit obtains the current work efficiency assessment result by integrating the shielding magnetic field compliance degree and the shielding signal compliance degree; The fusion model used by the comprehensive health analysis and decision-making unit to generate the health assessment index is an exponential function model. Specifically, it involves weighting and summing the historical loss assessment results and the current work efficiency assessment results, and then performing an exponential operation with the natural constant as the base to obtain the health assessment index.

[0008] In a preferred embodiment of the present invention, the operation and maintenance health assessment unit analyzes the health status of the shielding device using a shielding device health assessment formula. The shielding device integrates health and operation and maintenance health assessment formulas. , As a health assessment index, To assess the proportion and weight, It is a natural constant. To maintain operational health, , To shield the magnetic field compatibility, , The lowest electromagnetic intensity detected. The rate of decrease in magnetic field strength, and This is an electromagnetic reference value. As a correction factor, To mask signal compliance, , This refers to the duration of the shielding device's operation and the duration of its operation across each frequency band. The difference in the number of signals in the j-th frequency band. The highest signal amplitude detected. , , and These are the reference values ​​for each frequency band. , , and For weighting percentage, This is a correction factor.

[0009] As a preferred embodiment of the present invention, the examination room signal detection module includes a signal acquisition and preprocessing unit, a background signal learning and cancellation unit, a real-time signal feature extraction unit, and a detection information generation and reporting unit. The signal acquisition and preprocessing unit is used to perform wide-band, continuous radio signal scanning and acquisition in the examination room, convert the received analog radio frequency signals into digital signals, and perform preliminary filtering and gain control. The background signal learning and cancellation unit is connected to the signal acquisition and preprocessing unit. The background signal learning and cancellation unit is used to learn and establish a background signal feature model of the electromagnetic environment of the examination room before the examination or during non-examination periods. During the examination, the real-time acquired spectrum data is compared with the background signal feature model, and signals that match the background features are filtered out by a preset algorithm, thereby highlighting new and abnormal potential cheating signals. The real-time signal feature extraction unit is connected to the background signal learning and cancellation unit. The real-time signal feature extraction unit is used to analyze the abnormal signal pulses or communication links that have been screened out after background cancellation, and extract their feature parameters in multiple dimensions in the time domain, frequency domain, modulation domain and spatial domain in parallel. The detection information generation and reporting unit is connected to the real-time signal feature extraction unit. The detection information generation and reporting unit is used to format and encapsulate the extracted multi-dimensional feature parameters to generate a standard data packet containing signal identification, feature vector, timestamp and preliminary risk assessment indication, and report it to the detection signal parsing module in real time through the network communication interface. The signal acquisition and preprocessing unit specifically includes a wideband radio frequency front-end, a programmable gain control link, an analog-to-digital conversion subunit, and a field-programmable gate array preprocessing subunit. The wideband RF front-end is used to receive RF signals in the 70MHz to 6GHz frequency band. The programmable gain control link is used to dynamically adjust the attenuation and amplification based on the power of the input signal to prevent the analog-to-digital converter from saturating and to ensure sensitivity to small signals. The analog-to-digital conversion subunit is used to convert the conditioned analog signal into a high-sampling-rate digital signal. The field-programmable gate array preprocessing subunit is implemented based on FPGA and is used to perform digital down-conversion, channelization and preliminary filtering on the digital signal. The programmable gain control link adopts a structure of alternating cascade of multi-stage digitally controlled attenuators and low-noise amplifiers, and integrates a temperature-compensated attenuator to maintain gain stability over a wide temperature range. The analog-to-digital conversion module uses multiple ADC chips to operate in a time-interleaved sampling mode to achieve high sampling rate acquisition of ultra-wideband signals.

[0010] As a preferred embodiment of the present invention, the signal feature deep extraction and clustering unit is used to perform deep analysis on the raw radio signal data reported by the examination room signal detection module, extract multi-dimensional feature vectors of the signal in the time domain, frequency domain, modulation domain, protocol domain and behavior domain, and use historical data to establish a feature model library of background signals and typical cheating signals through unsupervised learning for real-time signal comparison and classification. The cheating risk quantification assessment unit is connected to the signal feature deep extraction and clustering unit. The cheating risk quantification assessment unit is used to construct a comprehensive assessment model based on the extracted signal feature vector and its comparison results with the feature model library, and to calculate the comprehensive cheating risk index of each detected signal or each examination room. This index is used to quantify the probability that the signal is a cheating signal and its potential degree of harm. The multi-source information fusion and decision-making unit is connected to the cheating risk quantification assessment unit. The multi-source information fusion and decision-making unit is used to receive the comprehensive cheating risk index, and combine it with static information of the examination room, dynamic monitoring information and preset control strategies to perform information fusion and decision-making judgment, determine the risk level and generate hierarchical control instructions. The control instructions include at least log recording, personnel prompts, on-site early warning and linkage enhancement shielding. The signal feature deep extraction and clustering unit specifically includes a feature extraction subunit, a deep clustering subunit, and a model library update subunit; The feature extraction subunit is used to extract pulse descriptors from the signal. The pulse descriptors include the signal's arrival time, carrier frequency, pulse width, angle of arrival, and pulse amplitude features. The deep clustering subunit is constructed based on a deep variational autoencoder. The deep clustering subunit is used to perform unsupervised clustering analysis on the multi-dimensional feature vector. The deep variational autoencoder includes an encoder, a reconstruction layer and a decoder. The encoder maps the input features to the latent space and constrains the distribution of the latent space by the reconstruction error and KL divergence loss, so as to learn the essential feature representation of the signal and automatically complete the clustering. The model library update subunit is used to dynamically add newly discovered anomalous signal clusters with stable characteristics in cluster analysis to the typical cheating signal feature model library after automatic confirmation. The deep clustering module uses a sequence neural network model based on an attention mechanism as a feature extractor. The sequence neural network model performs context awareness and fusion of the sequence features of the signal through a one-dimensional convolutional neural network and an attention mechanism, maps the original features to a highly separable metric space, and then uses a distance metric based on the inner product for clustering and sorting. In the aforementioned cheating risk quantification assessment unit, the formula for calculating the comprehensive cheating risk index is as follows: ,in Let M be the overall risk value of the i-th examination room, and M be the number of evaluation factors. Let m be the weight of the m-th factor. Let be the score of the i-th evaluation object on the m-th factor; As a preferred embodiment of the present invention, the multi-source information fusion and decision-making unit adopts a dynamic threshold hierarchical decision-making mechanism, specifically as follows: Set a safety threshold Level 1 threshold and secondary threshold The calculated comprehensive risk value The result is compared with the dynamic threshold, and hierarchical decision-making logic is executed based on the comparison result: like Deemed safe, only for record-keeping; like If the risk is determined to be low, it will trigger log recording and may send a prompt to the proctor; like If the risk level is determined to be medium, an on-site audio-visual warning will be triggered, and invigilators will be notified to conduct focused inspections. like If the risk level is identified as high, the detailed information should be immediately reported to the feedback warning terminal, and an instruction should be sent to the examination room shielding and analysis control module to request enhanced shielding of specific frequency bands or areas.

[0011] As a preferred embodiment of the present invention, the examination room shielding analysis control module includes an appropriate power and interference assessment unit, a multi-source risk fusion and decision-making unit, and a dynamic power calculation and instruction generation unit. The appropriate power and interference assessment unit is used to calculate the shielding power threshold based on the physical environment parameters and surrounding communication environment parameters of the target examination room, and to quantitatively assess the degree of electromagnetic interference caused to the public communication network outside the examination room under the current working state of the shielding device. The multi-source risk fusion and decision-making unit is connected to the appropriate power and interference assessment unit. The multi-source risk fusion and decision-making unit is used to receive and fuse multi-source risk signals from the health status of the shielding device, the risk of cheating before the exam, and the risk of cheating during the exam. Based on the fusion results, it intelligently determines whether the corresponding exam room needs power regulation and the urgency level of regulation. The dynamic power calculation and command generation unit is connected to the appropriate power and interference assessment unit and the multi-source risk fusion and decision unit, respectively. The dynamic power calculation and command generation unit is used to analyze the precise target shielding power for the examination room that is determined to need to be regulated, based on the appropriate shielding power threshold, the quantified interference level and the multi-source fusion risk level, through a preset dynamic calculation model, and generate a power regulation command that can be issued to the corresponding examination room shielding device for execution. The appropriate power and interference assessment unit specifically includes a power threshold calculation subunit and an interference assessment subunit. The power threshold calculation subunit is used to calculate the minimum effective power required to cover the examination room based on the spatial volume of the examination room, the attenuation coefficient of the building material, and the relative distance to the surrounding base stations, and uses it as the appropriate shielding power threshold. The interference assessment subunit is used to calculate the interference thermal noise ratio by monitoring the signal strength of a specific frequency band outside the boundary of the examination room, so as to quantify the interference level caused by the signal emitted by the shielding device to the uplink of the surrounding base stations. The interference assessment module uses a distributed sensor network for data collection, deploys multiple monitoring nodes around the examination room to simultaneously collect signal strength from different directions, and synthesizes a comprehensive external interference assessment index through a weighted average fusion algorithm to improve the spatial representativeness and accuracy of the assessment. The multi-source risk fusion and decision-making unit adopts a multi-modal feature fusion method based on attention mechanism. This method uses the shielding device health assessment index, the pre-exam detection anomaly assessment index, the in-exam detection anomaly assessment index, and the static risk factor from the exam room information extraction module as multi-source input features. It extracts deep features of each source data through an optimized convolutional neural network and uses the attention mechanism to dynamically calculate the weight of each feature vector in this decision, thereby achieving context-aware feature fusion. In the dynamic power calculation and instruction generation unit, the preset dynamic calculation model is an exponential amplification model, and its core calculation formula is: ,in For target shielding power, , and The evaluation criteria are divided into three main indices: interference shielding, pre-exam anomaly detection, and in-exam anomaly detection. , and These are the weighting percentages of the three major indices for the assessment of shielding and control measures. This is the exponential amplification factor.

[0012] As a preferred embodiment of the present invention, the feedback early warning terminal includes a core processing and intelligent analysis unit, an early warning execution and feedback unit, and a communication and coordination unit; The core processing and intelligent analysis unit is used to receive early warning trigger signals and related data from the examination room shielding analysis control module, the shielding device status analysis module and the detection signal analysis module, to preprocess the input data, extract features and perform fusion analysis, and to generate graded early warning decisions based on the preset risk assessment model. The early warning execution and feedback unit is connected to the core processing and intelligent analysis unit. The early warning execution and feedback unit is used to convert the hierarchical early warning decision into multimodal early warning information including at least sound, light and text and release it. At the same time, it provides a human-computer interaction interface to receive and upload external feedback information from examination room management personnel, forming a closed loop of early warning release and handling confirmation. The communication and collaboration unit is connected to the core processing and intelligent analysis unit and the early warning execution and feedback unit, respectively. The communication and collaboration unit is used to realize remote data transmission and control command reception between the terminal and the upper-level management and control platform and other examination room terminals, and supports collaborative sharing and linkage of early warning information among multiple terminals.

[0013] Compared with the prior art, the advantages of this invention are: This invention deeply integrates IoT, big data analytics, and intelligent decision-making technologies into the field of signal shielding. Through the shielding device status monitoring module and the shielding device status analysis module, each device is endowed with the ability to self-perceive and self-report its health status. Its workload, transmission efficiency, and suppression effect of each frequency band are monitored and evaluated in real time, realizing a leap from post-fault repair to pre-fault prediction. This fundamentally eliminates security vulnerabilities caused by equipment problems and greatly improves the inherent reliability of the entire shielding system. Secondly, the examination room shielding analysis and control module, as the intelligent hub, can comprehensively consider the static model of the examination room, real-time health data of the equipment, and dynamic environmental risks. It uses algorithms to perform multi-objective optimization calculations and dynamically outputs the most suitable target shielding power. It automatically adjusts according to the ambient temperature and human needs, ensuring that the shielding effect meets the standards in any complex examination room environment, while effectively suppressing excessive interference to public communication and legitimate electronic devices outside the examination room. This solves the inherent contradiction of fixed power and achieves precise, environmentally friendly, and humanized shielding.

[0014] This invention, through its signal analysis module and examination room shielding control module, can actively scan, receive, and intelligently analyze electromagnetic signals in the space. Its background signal learning and cancellation technology effectively filters out environmental noise, greatly enhancing the ability to detect concealed cheating signals. By performing deep feature extraction and risk model-based quantitative assessment of the signals, the system can not only detect signals but also understand their threat level and behavioral patterns, enabling tiered responses and targeted, enhanced, and precise blocking. Furthermore, all data throughout the entire process, from device status logs, original signal characteristics, risk assessment results to every power control command, is completely recorded, forming a data evidence chain. This provides irrefutable technical evidence for tracing cheating behavior and locating cheating devices after the exam, making the entire examination process auditable and reviewable, greatly enhancing the authority and persuasiveness of the exam's fairness. Attached Figure Description

[0015] Figure 1 This is a schematic diagram of a detection and control system for an AI intelligent signaling jammer used in an examination area according to the present invention. Detailed Implementation

[0016] 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 a part of the embodiments of the present invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0017] Please see Figure 1A detection and control system for an AI-powered intelligent signaling jammer in an examination room area includes: an examination room information extraction module, a jammer status monitoring module, a jammer status analysis module, an examination room signal detection module, a detection signal analysis module, an examination room jamming analysis and control module, and a feedback and early warning terminal. The exam room information extraction module is responsible for extracting and integrating basic data; The shielding device status monitoring module is used to monitor the working status of each shielding device in real time and calculate the health assessment index. When the index is lower than the set threshold, the system will determine that the device is abnormal. The shielding device status analysis module is used to perform in-depth analysis of the data collected by the shielding device status monitoring module, comprehensively evaluate the wear and tear and real-time working efficiency of the shielding device, and finally calculate a quantitative health assessment index. If the index is lower than the set threshold, the system will determine that the device is abnormal. The examination room signal detection module is used to continuously detect the radio signal environment within the examination room; The detection signal analysis module is used to analyze the raw data acquired by the detection module; The examination room shielding analysis and control module is used to integrate the analysis results of all the aforementioned modules, dynamically analyze and calculate the target shielding power of each examination room, and send instructions to the shielding device to adjust the power. The feedback and early warning terminal is used to immediately send an early warning instruction to the corresponding examination room administrator when the system detects an abnormal situation.

[0018] In a specific embodiment of the present invention, the examination room information extraction module integrates basic static data, and then the shielding device status monitoring module and the examination room signal detection module work in parallel to collect dynamic data on the device's own status and the electromagnetic environment of the examination room. Subsequently, the shielding device status analysis module and the detection signal analysis module perform in-depth analysis on the above data to quantitatively assess the device's health and the risk of cheating. The examination room shielding analysis and control module, as the decision-making center, integrates all analysis results, dynamically calculates the optimal target shielding power, and issues control commands. Finally, the feedback warning terminal ensures that any abnormality can be immediately communicated to the management personnel. The entire system realizes a paradigm shift from passive shielding to active, intelligent detection and control.

[0019] Specifically, the examination room information extraction module includes an examination room environment and configuration information extraction unit and a shielding device equipment file information extraction unit; The examination room environment and configuration information extraction unit is responsible for comprehensively collecting the inherent attributes of the examination room, the examination plan, and the information of the participants, so as to provide a basis for assessing the shielding difficulty, calculating the initial power, and managing personnel. The shielding device file information extraction unit is used to extract the shielding device's own history and status information; The information extracted by the examination room environment and configuration information extraction unit includes the examination room address, examination room volume, number of surrounding communication base stations, examination time period, number of invigilators, number of examinees, and facial images of each examinee. The information extracted by the jamming device equipment file information extraction unit includes the jamming device number, number of times the jamming device has been activated, cumulative activation time, and maintenance frequency.

[0020] In a specific embodiment of the present invention, the examination room environment and configuration information extraction unit is responsible for collecting the physical attributes, communication environment attributes, and examination information of the examination room. These are the basis for assessing the shielding difficulty, calculating the initial power, and verifying personnel identities. The shielding device file information extraction unit focuses on the full life cycle data management of the equipment, extracting and maintaining the serial number, historical activation count, cumulative working time, and maintenance records of each shielding device. This provides data traceability support for assessing the aging and wear of the equipment due to long-term use, realizing the structured and standardized collection and management of static information of the examination room and historical information of the equipment. It directly links information such as the volume of the examination room space and the density of surrounding base stations with the shielding power calculation, providing a scientific initial basis for subsequent precise power control, avoiding the blind setting of power, and establishing a complete electronic file of the equipment. This allows the equipment maintenance to shift from post-maintenance to state-based predictive maintenance. By analyzing data such as cumulative working time, it is possible to provide early warning of performance degradation that may be caused by aging, ensuring the overall health level of the equipment queue and improving the planning and foresight of the examination support work.

[0021] Specifically, the shielding device status monitoring module includes an instruction control and execution unit, a multi-parameter data acquisition unit, a data preprocessing and caching unit, and a preliminary status feedback and communication interface unit; The instruction control and execution unit is used to receive instructions from the examination room shielding analysis control module and send an open or close signal to the designated examination room shielding device; The multi-parameter data acquisition unit is used to collect operation and maintenance status information in real time. The operation and maintenance status information includes electromagnetic field strength, signal parameters of each shielded frequency band, number of signals and signal amplitude. The data preprocessing and caching unit is used to perform preliminary processing on the raw data acquired by the multi-parameter data acquisition unit. The preliminary processing includes data cleaning, formatting, and timestamp marking, and is organized and temporarily stored according to the structure of examination room, monitoring time point, and frequency band. The preliminary status feedback and communication interface unit is used to package the pre-processed operation and maintenance status information and transmit it to the shielding device status analysis module in real time through the system internal bus or network interface. At the same time, the preliminary status feedback and communication interface unit also receives query instructions from the shielding device status analysis module or the examination room shielding analysis control module and returns real-time data.

[0022] In a specific embodiment of the present invention, equipment status monitoring is upgraded from manual spot checks to fully automatic, continuous sensing of all parameters. By collecting data such as electromagnetic field strength, which directly reflect the shielding effect, in real time, it provides first-hand evidence for judging whether the equipment is actually working and how well it is working. This solves the problem of traditional methods that make it difficult to verify the effectiveness of each piece of equipment. The standardized data preprocessing and communication interface design ensures the compatibility and comparability of data from different examination rooms and different models of equipment, laying the foundation for unified health analysis at the upper level. This design also greatly reduces the workload of on-site technicians, freeing them from tedious equipment inspections and allowing them to focus on handling abnormal situations prompted by the system.

[0023] Specifically, the shielding device status analysis module includes a usage health assessment unit, an operation and maintenance health assessment unit, and a comprehensive health analysis and decision-making unit; The health assessment unit is used to assess the historical wear and tear or aging of the shielding device due to long-term use. The operation and maintenance health assessment unit is used to evaluate the real-time working efficiency of the shielding device after it is activated. At the same time, it directly analyzes the electromagnetic field effect generated by the device and its ability to suppress target signals, which is the key to determining whether it is working normally at present. The integrated health analysis and decision-making unit is used to combine usage health and operational health to generate a unified and comparable health assessment index, and make clear diagnostic decisions based on the index to trigger subsequent processes.

[0024] In a specific embodiment of the present invention, a hierarchical analytical model is introduced to achieve multi-dimensional and refined diagnosis of the health status of the signal jammer. The assessment of usage health and operational health is separated, enabling the detection of both potential chronic faults caused by long-term use and sudden acute faults that occur after startup. This provides more comprehensive diagnostic coverage and generates a unified health assessment index, allowing the status of hundreds or thousands of devices to be quantified, compared, and ranked. This facilitates management personnel in prioritizing the devices with the worst health, optimizing resource allocation. Based on data-driven predictive maintenance capabilities, device faults can be eliminated in their early stages, fundamentally preventing the significant risk of signal jamming failure in the examination room due to the signal jammer's own malfunction, and greatly improving the reliability of the examination security system.

[0025] Specifically, the operation and maintenance health assessment unit analyzes the health status of the shielding device using a shielding device health assessment formula, which is as follows: , As a health assessment index, To assess the proportion and weight, It is a natural constant. To maintain operational health, , To shield the magnetic field compatibility, , The lowest electromagnetic intensity detected. The rate of decrease in magnetic field strength, and This is an electromagnetic reference value. As a correction factor, To mask signal compliance, , This refers to the duration of the shielding device's operation and the duration of its operation across each frequency band. The difference in the number of signals in the j-th frequency band. The highest signal amplitude detected. , , and These are the reference values ​​for each frequency band. , , and For weighting percentage, This is a correction factor.

[0026] In a specific embodiment of the present invention, the algorithm transforms the originally vague concept of working status into precise and calculable indicators, realizing the objectification and scientification of equipment performance evaluation. Through multi-parameter fusion analysis, the algorithm can sensitively identify various fault modes. Insufficient electromagnetic field strength may indicate a power amplifier fault, while excessive residual signal may indicate frequency synthesizer inaccuracy or antenna fault. This enables the system not only to determine whether the equipment is powered on, but also to accurately determine whether it is working effectively and on which frequency bands its performance is compromised. This provides a direct basis for subsequent targeted maintenance and also provides underlying data support for dynamically adjusting the shielding power of different frequency bands, achieving truly precise control.

[0027] Specifically, the examination room signal detection module includes a signal acquisition and preprocessing unit, a background signal learning and cancellation unit, a real-time signal feature extraction unit, and a detection information generation and reporting unit; The signal acquisition and preprocessing unit is used to automatically and continuously scan and receive wireless communication signals in the examination environment, covering various frequency bands from common public mobile communications to those used by potential cheating devices, and converting analog radio frequency signals into digital signals that can be used for subsequent analysis. The background signal learning and cancellation unit is used to distinguish between background noise in the examination room and abnormal signals that occur during the examination. The signal filtering method used by the background signal learning and cancellation unit is to input real-time spectrum data into a trained autoencoder or classifier and directly determine whether it deviates from the normal background feature pattern. The real-time signal feature extraction unit is used to analyze the selected signal pulses or communication links and extract their multi-dimensional features in the time domain, frequency domain, information domain, and spatial domain. The features extracted by the real-time signal feature extraction unit include time domain features, frequency domain features, information domain features, and spatial domain features. Time domain features are used to analyze the occurrence time pattern of the signal, frequency domain features are used to analyze the frequency attributes of the signal, including determining its center frequency, signal bandwidth, and modulation method identification, information domain features are used to demodulate the signal and restore its transmitted information content, and spatial domain features are used to estimate the approximate direction or distance of the signal radiation source. The detection information generation and reporting unit is used to summarize and format real-time detection results, generate standardized detection information data packets, and report them to the detection signal analysis module through the network communication module.

[0028] In specific embodiments of the present invention, the system is equipped with the ability to proactively detect cheating threats, representing a generational difference from traditional systems that can only passively block cheating. Background learning and AI filtering technologies greatly reduce interference from inherent environmental signals, significantly improving the sensitivity and accuracy of detecting weak and concealed cheating signals. Multi-dimensional feature extraction not only helps identify signal types but also provides clues for analyzing cheating behavior patterns. This allows the exam security defense line to be moved significantly forward, from preventing signals from being transmitted to detecting cheating attempts. It provides invigilators with precise patrol clues and can be linked with the blocking system to achieve targeted strikes against specific cheating signals, constructing a complete closed loop of detection and blocking.

[0029] Specifically, the detection signal analysis module includes a signal feature deep extraction and clustering unit, a cheating risk quantification assessment unit, and a multi-source information fusion and decision-making unit; The signal feature deep extraction and clustering unit is used to perform refined processing and pattern learning on the reported raw detection data. The signal feature deep extraction and clustering unit performs deep analysis on the suspicious signals reported by the detection information generation and reporting unit, extracting their multi-dimensional feature vectors in the time domain, frequency domain, modulation domain, protocol domain, and behavior domain. At the same time, using historical data, it establishes a feature model library of normal background signals and typical cheating signals through unsupervised learning for subsequent real-time comparison and classification. The cheating risk quantification assessment unit is used to convert signal characteristics into measurable risk values. Based on the extracted signal characteristics and their comparison with the feature model library, the unit constructs a comprehensive assessment model to calculate the comprehensive cheating risk index for each detected signal or each examination room. This index should reflect the probability that the signal is a cheating signal and its potential harm. The calculation formula for the cheating risk quantification assessment unit is as follows: ,in Let M be the overall risk value of the i-th examination room, and M be the number of evaluation factors. Let m be the weight of the m-th factor. Let be the score of the i-th evaluation object on the m-th factor; The multi-source information fusion and decision-making unit is used to integrate all analysis results, make a final judgment and trigger corresponding actions. The multi-source information fusion and decision-making unit receives the output from the risk assessment unit, combines static information, dynamic information and preset strategies in the examination room, performs information fusion and decision-making, determines the risk level and generates clear control instructions. The multi-source information fusion and decision-making unit will calculate the comprehensive risk value. With dynamically adjusted risk level thresholds , and Comparison, , and These are the safety threshold, the first-level threshold, and the second-level threshold, respectively, and the decision logic is as follows: like Deemed safe, only for record-keeping; like If the risk is determined to be low, it will trigger log recording and may send a prompt to the proctor; like If the risk level is determined to be medium, an on-site audio-visual warning will be triggered, and invigilators will be notified to conduct focused inspections. like If the risk level is identified as high, the detailed information should be immediately reported to the feedback warning terminal, and an instruction should be sent to the examination room shielding and analysis control module to request enhanced shielding of specific frequency bands or areas.

[0030] In specific embodiments of the present invention, a leap from signal recognition to risk decision-making is achieved. By establishing a feature model library and a quantitative evaluation model, the system can perform analogical analysis and risk assessment on unknown new cheating signals, possessing a certain degree of adaptability and evolution. The dynamic hierarchical decision-making logic makes the control measures highly flexible and targeted, avoiding a one-size-fits-all overreaction. This ensures that while effectively combating high-tech cheating, it can minimize false interference with normal electronic devices in the examination room and optimize system resource allocation, accurately targeting the strongest shielding power to the highest-risk frequency bands or time periods, achieving the best balance between security and efficiency, and conforming to the modern concept of intelligent and refined management.

[0031] Specifically, the examination room shielding analysis and control module includes an appropriate power and interference assessment unit, a multi-source risk fusion and decision-making unit, and a dynamic power calculation and instruction generation unit; The appropriate power and interference assessment unit is used to establish the initial position of power adjustment and quantify the external impact of the shielding behavior itself. Based on the basic physical environment of the examination room and the surrounding communication conditions, the appropriate power and interference assessment unit calculates a theoretical appropriate shielding power threshold as a benchmark reference for power adjustment, and at the same time quantifies and assesses the degree of electromagnetic interference caused to the outside of the examination room under the current working state of the shielding device. The multi-source risk fusion and decision-making unit is used to fuse multiple risk signals from different modules to determine whether power regulation is required. The dynamic power calculation and instruction generation unit is used to calculate the precise target power for objects marked as controlled examination rooms and generate executable control instructions.

[0032] In specific embodiments of the present invention, dynamic and precise control of shielding power is achieved. By introducing appropriate power thresholds and interference assessments, the system ensures that power adjustments are always made within a safe and effective range. By integrating equipment risks and environmental risks, the system can make intelligent decisions. When the health of a certain examination room's equipment is acceptable but the risk of cheating suddenly increases, the system instructs to appropriately increase the power. When the equipment's own performance declines, it may trigger an early warning to replace the equipment instead of blindly increasing the power. This intelligent decision-making based on multi-source information fusion makes the shielding system an organic whole that adapts to changes in the environment and threats, rather than a bunch of isolated machines.

[0033] Specifically, the feedback and early warning terminal includes a core processing and intelligent analysis unit, an early warning execution and feedback unit, and a communication and coordination unit; The core processing and intelligent analysis unit is used to preprocess the input digital signals, extract features, assess the state, and predict risks, and generate early warning decisions based on preset models or algorithms. The early warning execution and feedback unit is used to transform analytical decisions into early warning information that can be perceived by humans and to receive external feedback. The early warning execution and feedback unit releases early warning information in multiple forms such as sound, light, electricity, and text, and provides a human-computer interaction interface to receive and upload feedback information from users or administrators. The communication and coordination unit is responsible for remotely transmitting terminal data, receiving control commands, and supporting collaborative work among multiple terminals within the system.

[0034] In a specific embodiment of the present invention, the terminal design realizes closed-loop management of early warning information from its arrival to effective handling. The intelligent analysis function improves the accuracy and operability of early warnings, reduces the interference of false alarms on management personnel, and the multimodal early warning release method ensures that information is delivered. Whether the management personnel are in the monitoring center or on the way to the examination, they can know the key situation in time. The feedback mechanism provides that management actions can be recorded and traced by the system, forming a complete chain of responsibility, which is convenient for post-event review and analysis. The communication and collaboration capabilities support the cross-regional and cross-level sharing and collaborative handling of early warning information in large examination sites or multi-examination site linkage scenarios, which greatly improves the overall scheduling and emergency response efficiency of the examination command center and integrates the scattered examination sites into a unified command and rapid response intelligent prevention and control network.

[0035] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A detection and control system for an AI intelligent signaling jammer in an examination room area, characterized in that, include: The system includes: an examination room information extraction module, a jamming device status monitoring module, a jamming device status analysis module, an examination room signal detection module, a detection signal analysis module, an examination room jamming analysis control module, and a feedback and early warning terminal. The examination room information extraction module is used to extract environmental configuration information and shielding device file information of each examination room in the target examination site. The shielding device status monitoring module is used to receive control commands and send on or off signals to the shielding devices in the designated examination room, and to collect the operation and maintenance status information of each shielding device in real time and automatically. The shielding device status analysis module is used to perform hierarchical analysis and quantitative evaluation of the health status of the shielding device; The examination room signal detection module is used to continuously detect the radio signal environment in the examination room. The detection signal parsing module is used to parse the raw data acquired by the examination room signal detection module; The examination room shielding analysis and control module is used to make intelligent decisions based on multi-source information fusion, dynamically analyze and calculate the target shielding power of each examination room, and send instructions to the shielding device to adjust the power. The feedback and early warning terminal is used to receive decisions from the examination room shielding and analysis control module and execute early warnings, while realizing closed-loop management and collaborative handling of early warning information.

2. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The examination room information extraction module includes an examination room environment and configuration information extraction unit and a shielding device file information extraction unit. The examination room environment and configuration information extraction unit is used to collect the spatial volume of the examination room, the attenuation coefficient of the building material and the distribution information of the surrounding communication base stations. The spatial volume and the attenuation coefficient of the building material are used by the examination room shielding analysis and control module to calculate the theoretically suitable shielding power threshold. The shielding device file information extraction unit is used to extract the shielding device's own history and status information.

3. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The shielding device status monitoring module includes an instruction control and execution unit, a multi-parameter data acquisition unit, a data preprocessing and caching unit, and a preliminary status feedback and communication interface unit. The instruction control and execution unit is used to receive remote instructions from the examination room shielding analysis control module and send control signals to the designated examination room signal jammer to turn it on or off. The multi-parameter data acquisition unit is used to collect the operation and maintenance status information of the target examination room signal jammer in real time. The operation and maintenance status information includes the electromagnetic field strength reflecting the shielding effect, the signal parameters of each shielding frequency band, the number of signals, and the signal amplitude. The data preprocessing and caching unit is used to perform preliminary processing on the raw operation and maintenance status data acquired by the multi-parameter data acquisition unit. The preliminary processing includes data cleaning, format unification and timestamp marking, and is organized and temporarily stored according to the three-dimensional structure of examination room identification, monitoring time point and signal frequency band. The preliminary status feedback and communication interface unit is used to package the pre-processed and structured operation and maintenance status information and transmit it to the backend shielding device status analysis module in real time through the network interface. At the same time, the preliminary status feedback and communication interface unit is also used to receive query instructions from the shielding device status analysis module or the examination room shielding analysis control module and return the corresponding real-time or historical data.

4. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The shielding device status analysis module includes a usage health assessment unit, an operation and maintenance health assessment unit, and a comprehensive health analysis and decision-making unit. The health assessment unit is used to assess the historical wear and tear or aging of the shielding device due to long-term use based on its historical usage data. The operation and maintenance health assessment unit is used to assess the current working efficiency of the shielding device based on the real-time monitoring data after it is activated. The current working efficiency includes the electromagnetic field effect generated by the device and its ability to suppress target communication signals. The integrated health analysis and decision-making unit is connected to the usage health assessment unit and the operation and maintenance health assessment unit. The integrated health analysis and decision-making unit is used to integrate the historical wear and tear assessment results with the current work efficiency assessment results to generate a unified, quantifiable and comparable health assessment index. Based on the comparison result of the health assessment index and the preset threshold, it triggers the diagnostic decision and subsequent control process for the shielding device. The operation and maintenance health assessment unit includes a shielded magnetic field compliance calculation subunit and a shielded signal compliance calculation subunit; The shielding magnetic field compliance calculation subunit is used to calculate a compliance index reflecting the magnetic field generation capability of the shielding device based on real-time monitored electromagnetic field strength data. The shielding signal compliance calculation subunit is used to calculate the compliance index reflecting the suppression effect of the shielding device on the target signal based on the signal parameters of each target shielding frequency band monitored in real time. The shielding magnetic field compliance calculation subunit locates the lowest electromagnetic field strength and the statistically obtained magnetic field strength reduction rate from the real-time monitoring data, and compares them with the preset normal reference value. The operation and maintenance health assessment unit obtains the current work efficiency assessment result by integrating the shielding magnetic field compliance degree and the shielding signal compliance degree; The fusion model used by the comprehensive health analysis and decision-making unit to generate the health assessment index is an exponential function model. Specifically, it involves weighting and summing the historical loss assessment results and the current work efficiency assessment results, and then performing an exponential operation with the natural constant as the base to obtain the health assessment index.

5. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 4, characterized in that, The operation and maintenance health assessment unit analyzes the health status of the shielding device using a shielding device health assessment formula. The shielding device integrates health and operation and maintenance health assessment formulas. , As a health assessment index, To assess the proportion and weight, It is a natural constant. To maintain operational health, , To shield the magnetic field compatibility, , The lowest electromagnetic intensity detected. The rate of decrease in magnetic field strength, and This is an electromagnetic reference value. As a correction factor, To mask signal compliance, ,in, This refers to the duration of the shielding device's operation and the duration of its operation across each frequency band. The difference in the number of signals in the j-th frequency band. The highest signal amplitude detected. , , and These are the reference values ​​for each frequency band. , , and For weighting percentage, This is a correction factor.

6. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The examination room signal detection module includes a signal acquisition and preprocessing unit, a background signal learning and cancellation unit, a real-time signal feature extraction unit, and a detection information generation and reporting unit. The signal acquisition and preprocessing unit is used to perform wide-band, continuous radio signal scanning and acquisition in the examination room, convert the received analog radio frequency signals into digital signals, and perform preliminary filtering and gain control. The background signal learning and cancellation unit is connected to the signal acquisition and preprocessing unit. The background signal learning and cancellation unit is used to learn and establish a background signal feature model of the electromagnetic environment of the examination room before the examination or during non-examination periods. During the examination, the real-time acquired spectrum data is compared with the background signal feature model, and signals that match the background features are filtered out by a preset algorithm, thereby highlighting new and abnormal potential cheating signals. The real-time signal feature extraction unit is connected to the background signal learning and cancellation unit. The real-time signal feature extraction unit is used to analyze the abnormal signal pulses or communication links that have been screened out after background cancellation, and extract their feature parameters in multiple dimensions in the time domain, frequency domain, modulation domain and spatial domain in parallel. The detection information generation and reporting unit is connected to the real-time signal feature extraction unit. The detection information generation and reporting unit is used to format and encapsulate the extracted multi-dimensional feature parameters to generate a standard data packet containing signal identification, feature vector, timestamp and preliminary risk assessment indication, and report it to the detection signal parsing module in real time through the network communication interface. The signal acquisition and preprocessing unit specifically includes a wideband radio frequency front-end, a programmable gain control link, an analog-to-digital conversion subunit, and a field-programmable gate array preprocessing subunit. The wideband RF front-end is used to receive RF signals in the 70MHz to 6GHz frequency band. The programmable gain control link is used to dynamically adjust the attenuation and amplification based on the power of the input signal to prevent the analog-to-digital converter from saturating and to ensure sensitivity to small signals. The analog-to-digital conversion subunit is used to convert the conditioned analog signal into a high-sampling-rate digital signal. The field-programmable gate array preprocessing subunit is implemented based on FPGA and is used to perform digital down-conversion, channelization and preliminary filtering on the digital signal. The programmable gain control link adopts a structure of alternating cascade of multi-stage digitally controlled attenuators and low-noise amplifiers, and integrates a temperature-compensated attenuator to maintain gain stability over a wide temperature range. The analog-to-digital conversion module uses multiple ADC chips to operate in a time-interleaved sampling mode to achieve high sampling rate acquisition of ultra-wideband signals.

7. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The detection signal analysis module includes a signal feature deep extraction and clustering unit, a cheating risk quantitative assessment unit, and a multi-source information fusion and decision-making unit. The signal feature deep extraction and clustering unit is used to perform deep analysis on the raw radio signal data reported by the examination room signal detection module, extract multi-dimensional feature vectors of the signal in the time domain, frequency domain, modulation domain, protocol domain and behavior domain, and use historical data to build a feature model library of background signals and typical cheating signals through unsupervised learning for real-time signal comparison and classification. The cheating risk quantification assessment unit is connected to the signal feature deep extraction and clustering unit. The cheating risk quantification assessment unit is used to construct a comprehensive assessment model based on the extracted signal feature vector and its comparison results with the feature model library, and to calculate the comprehensive cheating risk index of each detected signal or each examination room. This index is used to quantify the probability that the signal is a cheating signal and its potential degree of harm. The multi-source information fusion and decision-making unit is connected to the cheating risk quantification assessment unit. The multi-source information fusion and decision-making unit is used to receive the comprehensive cheating risk index, and combine it with static information of the examination room, dynamic monitoring information and preset control strategies to perform information fusion and decision-making judgment, determine the risk level and generate hierarchical control instructions. The control instructions include at least log recording, personnel prompts, on-site early warning and linkage enhancement shielding. The signal feature deep extraction and clustering unit specifically includes a feature extraction subunit, a deep clustering subunit, and a model library update subunit; The feature extraction subunit is used to extract pulse descriptors from the signal. The pulse descriptors include the signal's arrival time, carrier frequency, pulse width, angle of arrival, and pulse amplitude features. The deep clustering subunit is constructed based on a deep variational autoencoder. The deep clustering subunit is used to perform unsupervised clustering analysis on the multi-dimensional feature vector. The deep variational autoencoder includes an encoder, a reconstruction layer and a decoder. The encoder maps the input features to the latent space and constrains the distribution of the latent space by the reconstruction error and KL divergence loss, so as to learn the essential feature representation of the signal and automatically complete the clustering. The model library update subunit is used to dynamically add newly discovered anomalous signal clusters with stable characteristics in cluster analysis to the typical cheating signal feature model library after automatic confirmation. The deep clustering module uses a sequence neural network model based on an attention mechanism as a feature extractor. The sequence neural network model performs context awareness and fusion of the sequence features of the signal through a one-dimensional convolutional neural network and an attention mechanism, maps the original features to a highly separable metric space, and then uses a distance metric based on the inner product for clustering and sorting. In the cheating risk quantification assessment unit, the formula for calculating the comprehensive cheating risk index is as follows: ,in Let M be the overall risk value of the i-th examination room, and M be the number of assessment factors. Let m be the weight of the m-th factor. Let be the score of the i-th evaluation object on the m-th factor.

8. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 7, characterized in that, The multi-source information fusion and decision-making unit adopts a dynamic threshold hierarchical decision-making mechanism, specifically: Set a safety threshold Level 1 threshold and secondary threshold The calculated comprehensive risk value The result is compared with the dynamic threshold, and hierarchical decision-making logic is executed based on the comparison result: like Deemed safe, only for record-keeping; like If the risk is determined to be low, it will trigger log recording and may send a prompt to the proctor; like If the risk level is determined to be medium, an on-site audio-visual warning will be triggered, and invigilators will be notified to conduct focused inspections. like If the risk level is identified as high, the detailed information should be immediately reported to the feedback warning terminal, and an instruction should be sent to the examination room shielding and analysis control module to request enhanced shielding of specific frequency bands or areas.

9. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The examination room shielding analysis and control module includes an appropriate power and interference assessment unit, a multi-source risk fusion and decision-making unit, and a dynamic power calculation and instruction generation unit. The appropriate power and interference assessment unit is used to calculate the shielding power threshold based on the physical environment parameters and surrounding communication environment parameters of the target examination room, and to quantitatively assess the degree of electromagnetic interference caused to the public communication network outside the examination room under the current working state of the shielding device. The multi-source risk fusion and decision-making unit is connected to the appropriate power and interference assessment unit. The multi-source risk fusion and decision-making unit is used to receive and fuse multi-source risk signals from the health status of the shielding device, the risk of cheating before the exam, and the risk of cheating during the exam. Based on the fusion results, it intelligently determines whether the corresponding exam room needs power regulation and the urgency level of regulation. The dynamic power calculation and command generation unit is connected to the appropriate power and interference assessment unit and the multi-source risk fusion and decision unit, respectively. The dynamic power calculation and command generation unit is used to analyze the precise target shielding power for the examination room that is determined to need to be regulated, based on the appropriate shielding power threshold, the quantified interference level and the multi-source fusion risk level, through a preset dynamic calculation model, and generate a power regulation command that can be issued to the corresponding examination room shielding device for execution. The appropriate power and interference assessment unit specifically includes a power threshold calculation subunit and an interference assessment subunit. The power threshold calculation subunit is used to calculate the minimum effective power required to cover the examination room based on the spatial volume of the examination room, the attenuation coefficient of the building material, and the relative distance to the surrounding base stations, and uses it as the appropriate shielding power threshold. The interference assessment subunit is used to calculate the interference thermal noise ratio by monitoring the signal strength of a specific frequency band outside the boundary of the examination room, so as to quantify the interference level caused by the signal emitted by the shielding device to the uplink of the surrounding base stations. The interference assessment module uses a distributed sensor network for data collection, deploys multiple monitoring nodes around the examination room to simultaneously collect signal strength from different directions, and synthesizes a comprehensive external interference assessment index through a weighted average fusion algorithm to improve the spatial representativeness and accuracy of the assessment. The multi-source risk fusion and decision-making unit adopts a multi-modal feature fusion method based on attention mechanism. This method uses the shielding device health assessment index, the pre-exam detection anomaly assessment index, the in-exam detection anomaly assessment index, and the static risk factor from the exam room information extraction module as multi-source input features. It extracts deep features of each source data through an optimized convolutional neural network and uses the attention mechanism to dynamically calculate the weight of each feature vector in this decision, thereby achieving context-aware feature fusion. In the dynamic power calculation and instruction generation unit, the preset dynamic calculation model is an exponential amplification model, and its core calculation formula is: ,in For target shielding power, , and The evaluation criteria are divided into three main indices: interference shielding, pre-exam anomaly detection, and in-exam anomaly detection. , and These are the weighting percentages of the three major indices for the assessment of shielding and control measures. This is the exponential amplification factor.

10. The detection and control system for an AI intelligent signaling jammer in an examination room area according to claim 1, characterized in that, The feedback and early warning terminal includes a core processing and intelligent analysis unit, an early warning execution and feedback unit, and a communication and coordination unit. The core processing and intelligent analysis unit is used to receive early warning trigger signals and related data from the examination room shielding analysis control module, the shielding device status analysis module and the detection signal analysis module, to preprocess the input data, extract features and perform fusion analysis, and to generate graded early warning decisions based on the preset risk assessment model. The early warning execution and feedback unit is connected to the core processing and intelligent analysis unit. The early warning execution and feedback unit is used to convert the hierarchical early warning decision into multimodal early warning information including at least sound, light and text and release it. At the same time, it provides a human-computer interaction interface to receive and upload external feedback information from examination room management personnel, forming a closed loop of early warning release and handling confirmation. The communication and collaboration unit is connected to the core processing and intelligent analysis unit and the early warning execution and feedback unit, respectively. The communication and collaboration unit is used to realize remote data transmission and control command reception between the terminal and the upper-level management and control platform and other examination room terminals, and supports collaborative sharing and linkage of early warning information among multiple terminals.

Images