A chip-based internet of things device dynamic signal control optimization system and method

Abstract

This invention discloses a chip-based dynamic signal control optimization system and method for IoT devices, relating to the field of chip control technology. During system operation, the data acquisition module monitors and records the original transmission signals and operational resource information in real time, providing basic data. The signal processing module filters, amplifies, and modulates the data to generate a precise signal feature vector X, enabling the communication processing module to effectively extract data frame, packet loss rate, and transmission / reception time information, thereby calculating the bit error rate P. error By using deep learning technology to establish a control operation resource evaluation model, the operation status evaluation coefficient Zxs is obtained and matched with the preset evaluation threshold P to formulate an optimization scheme and perform dynamic regulation, which improves the energy efficiency of chip operation and effectively reduces the bit error rate. It solves the shortcomings of traditional systems in performance monitoring, real-time adjustment and efficiency optimization, and realizes intelligent management and stable operation of chip control system.

Description

Technical Field

[0001] This invention relates to the field of chip control technology, specifically to a chip-based dynamic signal control optimization system and method for Internet of Things (IoT) devices. Background Technology

[0002] Currently, there are some significant shortcomings and deficiencies in the chip control systems of IoT devices. First, in terms of signal processing, input signals are easily affected by noise interference, leading to a decline in signal quality and thus affecting the normal operation of the device. Second, during data transmission, existing communication control algorithms often have a high bit error rate, affecting the accuracy and reliability of data transmission. In addition, IoT devices typically require low power consumption and high efficiency, while existing systems still have considerable room for improvement in energy efficiency optimization, resulting in shortened battery life and serious resource waste.

[0003] These shortcomings are mainly due to the following reasons: improper signal processing is because existing filtering and modulation algorithms fail to fully consider various noise sources and signal distortion factors, resulting in poor signal quality after processing; high bit error rate in data transmission stems from complex transmission environments and imperfect error correction mechanisms, which can lead to data loss or transmission errors, thereby affecting system reliability and user experience; and insufficient energy efficiency optimization is mainly due to the failure to fully consider resource utilization and power consumption management during system design, causing problems such as overheating, performance degradation, and rapid battery depletion during long-term operation. These abnormal effects not only affect the service life of the equipment but also increase maintenance costs and user dissatisfaction. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a chip-based dynamic signal control optimization system and method for Internet of Things (IoT) devices, which solves the problems mentioned in the background section.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solution: a chip-based dynamic signal control optimization system for Internet of Things devices, comprising a data acquisition module, a signal processing module, a communication processing module, a system processing module, and a monitoring feedback module;

[0006] The data acquisition module collects raw transmission signals in real time through sensors, and forms a data set by recording the power consumption and resource information used during operation and timestamp information.

[0007] The signal processing module preprocesses the acquired data set, including filtering, signal amplification and modulation, and simultaneously integrates the processed acquired data set into a signal feature vector X.

[0008] The communication processing module extracts features from the signal feature vector X, including data frames, packet loss rate, and transmission / reception time information, and calculates the bit error rate P. error ;

[0009] The system processing module uses deep learning technology to establish a control operation resource assessment model, and obtains the operation control energy efficiency index E by training and analyzing the control operation resource assessment model. total ;

[0010] The monitoring and feedback module will control the energy efficiency index E. total With bit error rate P error The system performs fitting to obtain the operation status evaluation coefficient Zxs, and matches it with the preset operation status evaluation threshold P to obtain the operation status efficiency evaluation scheme. Then, it performs specific execution and regulation according to the content of the operation status efficiency evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after regulation.

[0011] Preferably, the data acquisition module includes a signal recording unit and a resource monitoring unit;

[0012] The signal recording unit captures the transmitted signals in the system in real time by using sensors and signal sensing devices, synchronously receives and records signal data from the transmission channel, and then stores the captured raw signal data in time sequence to form a raw signal data group, including component temperature, acquisition timestamp, current value, voltage value, signal strength and signal frequency.

[0013] The sensors include current sensors, voltage sensors, temperature sensors, and signal strength sensors;

[0014] The resource monitoring unit monitors the power consumption in real time, including CPU utilization and memory usage, synchronously records timestamp information to mark the data acquisition time, assembles power consumption data, and combines it with the original signal data group to obtain the acquired data group.

[0015] Preferably, the signal processing module includes a signal preprocessing unit and a feature extraction unit;

[0016] The signal preprocessing unit preprocesses the acquired data set, including filtering, signal amplification, and modulation. Filtering includes using low-pass, high-pass, and band-pass filters to filter the original signal in the acquired data set, removing noise and interference. Signal amplification includes using operational amplifiers and low-noise amplifiers to amplify the filtered signal to adjust its intensity. Modulation includes using amplitude modulation and frequency modulation to modulate the amplified signal to conform to a preset transmission standard and format.

[0017] The feature extraction unit extracts feature parameters from the preprocessed signal to form a signal feature vector X, including Fourier transform extraction, time-frequency analysis extraction, and wavelet transform. The synchronously extracted acquisition data groups are integrated into the signal feature vector X, which includes operating power consumption, operating time, resource usage, bit error rate, signal strength, and signal frequency.

[0018] Preferably, the communication processing module includes a data frame unit;

[0019] The data frame unit extracts data frame information from the signal feature vector X, including the data frame, packet loss rate, data reception and transmission timestamps, and total number of data packets, and calculates the packet loss rate P. loss Data latency D delay and bit error rate P error ;

[0020] The packet loss rate P loss Obtained through the following calculation formula:

[0021]

[0022] In the formula, N lost The amount of data packets lost is represented by N, specifically calculated from the difference between the number of data packets recorded by the sender and the number reported by the receiver. sent Indicates the amount of data in the data packet sent;

[0023] The data latency D delay Obtained through the following calculation formula:

[0024]

[0025] In the formula, T total N represents the total delay time, specifically obtained by summing the transmission delay times of all received data packets recorded by the receiving end. received Indicates the number of data packets received;

[0026] The bit error rate P error Obtained through the following calculation formula:

[0027]

[0028] In the formula, N error This indicates the number of bits that were erroneous, specifically calculated by comparing the received data frame with a preset transmitted data frame; N total This represents the total number of bits, which is specifically obtained by recording and counting the total number of bits sent at the sending end.

[0029] Preferably, the system processing module includes a modeling unit;

[0030] The modeling unit uses deep learning technology to establish a control operation resource assessment model, and constructs an input layer, a hidden layer, and an output layer. The input layer receives the feature vector X and inputs the feature vector X into the control operation resource assessment model. The hidden layer performs nonlinear transformations and trains and analyzes the control operation resource assessment model. The output layer obtains the operation control energy efficiency index E. total The model parameters are iteratively adjusted using the loss function L(θ) and gradient descent algorithm.

[0031] Preferably, the input layer receives feature vector X, normalizes it, and extracts the total dimension d of feature vector X to obtain the received data of the input layer: normalized feature vector X. d ={X1,X2,X3,,,,,X d};

[0032] The hidden layer consists of several hidden layers, and the output is obtained through the following calculation steps:

[0033] First hidden layer output:

[0034] H1=σ*(W1*X d +b1);

[0035] Second hidden layer output:

[0036] H2 = σ*(W2*H1+b2);

[0037] Output of the p-th hidden layer:

[0038] H p =σ*(W p *H p-1 +b p );

[0039] In the formula, W1, W2 and W p The weight matrices represent the weight matrices of the first hidden layer, the second hidden layer, and the p-th hidden layer, respectively, with b1, b2, and b... p The bias vectors represent the bias vectors of the first hidden layer, the second hidden layer, and so on up to the p-th hidden layer. σ represents the activation function, including the ReLU function, H1, H2, and H... p These represent the outputs of the first hidden layer, the second hidden layer, and so on up to the p-th hidden layer, with the p-th hidden layer output H respectively. p Specifically, the energy efficiency index E of the output layer's operation control. total .

[0040] Preferably, the loss function L(θ) is obtained by the following formula:

[0041]

[0042] In the formula, L(θ) represents the loss function, specifically used to measure the preset operating control energy efficiency index and the operating control energy efficiency index E of the i-th sample. total,i The performance gap, E total,i This represents the operating control energy efficiency index of the i-th sample. This represents the preset operating control energy efficiency index, and N represents the total number of samples.

[0043] The gradient descent algorithm is adjusted using the following calculation formula:

[0044]

[0045] In the formula, θ represents the set of training parameters for controlling the operation of the resource assessment model, and α represents the learning rate, which is used to control the step size for each parameter update. This represents the gradient of the loss function with respect to the control and runtime resource evaluation model, including the partial derivatives of each parameter;

[0046] The training parameter set θ includes the weight matrix W1 (first hidden), the weight matrix W2 (second hidden), and W... p The weight matrix of the p-th hidden layer, b1 the bias vector of the first hidden layer, b2 the bias vector of the second hidden layer, and b p The bias vector of the p-th hidden layer.

[0047] Preferably, the monitoring feedback module includes a matching unit and a control unit;

[0048] The matching unit matches preset relevant information with the required comparison value to determine the operating control energy efficiency index E. total With bit error rate P error The system performs fitting to obtain the operating status evaluation coefficient Zxs, and matches it with the preset operating status evaluation threshold P to obtain the operating status performance evaluation scheme.

[0049] The operating status evaluation coefficient Zxs is obtained through the following calculation formula:

[0050]

[0051] In the formula, z1 and z2 represent the operating control energy efficiency index E, respectively. total With bit error rate P error The weighting coefficients, where U represents the correction constant;

[0052] The control unit performs specific execution and control according to the content of the operation status performance evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after control.

[0053] Preferably, the operational performance evaluation scheme is obtained through the following matching method:

[0054] If the operational status evaluation coefficient Zxs≤P, a qualified operational status performance evaluation result is obtained.

[0055] If the operating status evaluation coefficient Zxs > P, and the operating status performance evaluation result is unqualified, the operating parameters, including power management parameters, data transmission parameters, cache and storage parameters, task scheduling parameters, and temperature management parameters, will be adjusted.

[0056] A chip-based method for dynamic signal control optimization in IoT devices includes the following steps:

[0057] Step 1: The data acquisition module collects raw transmission signals in real time through sensors, and forms a data set by recording the power consumption, resource usage information and timestamp information during operation;

[0058] Step 2: The signal processing module preprocesses the acquired data set, including filtering, signal amplification and modulation, and simultaneously integrates the processed acquired data set into a signal feature vector X;

[0059] Step 3: The communication processing module extracts features from the signal feature vector X, including data frames, packet loss rate, and transmission / reception time information, and calculates the bit error rate P. error ;

[0060] Step 4: The system processing module uses deep learning technology to establish a control operation resource assessment model, and obtains the operation control energy efficiency index E by training and analyzing the control operation resource assessment model. total ;

[0061] Step 5: The monitoring and feedback module will control the energy efficiency index E. total With bit error rate P error The system performs fitting to obtain the operation status evaluation coefficient Zxs, and matches it with the preset operation status evaluation threshold P to obtain the operation status efficiency evaluation scheme. Then, it performs specific execution and regulation according to the content of the operation status efficiency evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after regulation.

[0062] This invention provides a chip-based dynamic signal control optimization system and method for Internet of Things (IoT) devices, which has the following advantages:

[0063] (1) During system operation, the system can accurately record the original transmission signals and operating resource information through real-time monitoring by the data acquisition module, providing rich basic data for subsequent processing. The signal processing module filters, amplifies, and modulates the data to generate an accurate signal feature vector X, enabling the communication processing module to effectively extract data frame, packet loss rate, and transmission / reception time information, thereby calculating the bit error rate P. error The system processing module uses deep learning technology to establish a control operation resource evaluation model, accurately obtains the operation control energy efficiency index, and provides a scientific basis for system performance evaluation. The monitoring and feedback module comprehensively analyzes the energy efficiency index and the bit error rate to obtain the operation status evaluation coefficient Zxs, and matches it with the preset evaluation threshold P to formulate optimization schemes and perform dynamic regulation, thereby improving the energy efficiency of chip operation and effectively reducing the bit error rate. This solves the shortcomings of traditional systems in performance monitoring, real-time adjustment and efficiency optimization, and realizes intelligent management and stable operation of the chip control system.

[0064] (2) The input layer receives the feature vector X and inputs the feature vector X into the control operation resource assessment model. The hidden layer performs nonlinear transformation and trains and analyzes the control operation resource assessment model. The output layer obtains the operation control energy efficiency index E. total The model parameters are iteratively adjusted using the loss function L(θ) and gradient descent algorithm. By utilizing layer-by-layer nonlinear transformation and iterative optimization algorithm, the energy efficiency index of the operation control is accurately calculated. The application of gradient descent algorithm ensures the precise adjustment of model parameters, thereby continuously improving the energy efficiency and performance of the system.

[0065] (3) By controlling the energy efficiency index E during operation total With bit error rate P error Fitting is performed to obtain the operating state evaluation coefficient Zxs, which is then matched with the preset operating state evaluation threshold P to obtain the operating state performance evaluation scheme. This enables efficient monitoring and optimization of the chip's operating state. The evaluation mechanism ensures that the system can promptly identify deficiencies in operation and formulate specific optimization measures based on the evaluation results. This helps to optimize the chip's energy efficiency and processing capabilities, and can also effectively reduce the bit error rate and improve system stability. Thus, it realizes intelligent management and dynamic optimization of the chip control system. Through this precise feedback and adjustment mechanism, the system can continuously improve performance and solve the shortcomings of traditional systems in real-time control and optimization. Attached Figure Description

[0066] Figure 1 This is a schematic diagram of the block diagram of a chip-based IoT device dynamic signal control optimization system according to the present invention.

[0067] Figure 2This is a schematic diagram illustrating the steps of a chip-based dynamic signal control optimization method for IoT devices according to the present invention. Detailed Implementation

[0068] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. 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.

[0069] Example 1

[0070] This invention provides a chip-based dynamic signal control optimization system for Internet of Things (IoT) devices. Please refer to [link / reference]. Figure 1 It includes a data acquisition module, a signal processing module, a communication processing module, a system processing module, and a monitoring and feedback module;

[0071] The data acquisition module collects raw transmission signals in real time through sensors, and forms a data set by recording the power consumption and resource information used during operation and timestamp information.

[0072] The signal processing module preprocesses the acquired data set, including filtering, signal amplification and modulation, and simultaneously integrates the processed acquired data set into a signal feature vector X.

[0073] The communication processing module extracts features from the signal feature vector X, including data frames, packet loss rate, and transmission / reception time information, and calculates the bit error rate P. error ;

[0074] The system processing module uses deep learning technology to establish a control operation resource assessment model, and obtains the operation control energy efficiency index E by training and analyzing the control operation resource assessment model. total ;

[0075] The monitoring and feedback module will control the energy efficiency index E. total With bit error rate P error The system performs fitting to obtain the operation status evaluation coefficient Zxs, and matches it with the preset operation status evaluation threshold P to obtain the operation status efficiency evaluation scheme. Then, it performs specific execution and regulation according to the content of the operation status efficiency evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after regulation.

[0076] In this embodiment, through real-time monitoring by the data acquisition module, the system can accurately record the original transmission signals and operational resource information, providing rich basic data for subsequent processing. The signal processing module filters, amplifies, and modulates the data to generate an accurate signal feature vector X, enabling the communication processing module to effectively extract data frame, packet loss rate, and transmission / reception time information, thereby calculating the bit error rate P. error The system processing module uses deep learning technology to establish a control operation resource evaluation model, accurately obtains the operation control energy efficiency index, and provides a scientific basis for system performance evaluation. The monitoring and feedback module comprehensively analyzes the energy efficiency index and the bit error rate to obtain the operation status evaluation coefficient Zxs, and matches it with the preset evaluation threshold P to formulate optimization schemes and perform dynamic regulation, thereby improving the energy efficiency of chip operation and effectively reducing the bit error rate. This solves the shortcomings of traditional systems in performance monitoring, real-time adjustment and efficiency optimization, and realizes intelligent management and stable operation of the chip control system.

[0077] Example 2

[0078] This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically: the data acquisition module includes a signal recording unit and a resource monitoring unit;

[0079] The signal recording unit captures the transmitted signals in the system in real time by using sensors and signal sensing devices, synchronously receives and records signal data from the transmission channel, and then stores the captured raw signal data in time sequence to form a raw signal data group, including component temperature, acquisition timestamp, current value, voltage value, signal strength and signal frequency.

[0080] The sensors include current sensors, voltage sensors, temperature sensors, and signal strength sensors;

[0081] The resource monitoring unit monitors the power consumption in real time, including CPU utilization and memory usage, and synchronously records timestamp information to mark the data acquisition time. This data is then combined with the original signal data set to obtain the acquired data set. The signal processing module includes a signal preprocessing unit and a feature extraction unit.

[0082] The signal preprocessing unit preprocesses the acquired data set, including filtering, signal amplification, and modulation. Filtering includes using low-pass, high-pass, and band-pass filters to filter the original signal in the acquired data set, removing noise and interference. Signal amplification includes using operational amplifiers and low-noise amplifiers to amplify the filtered signal to adjust its intensity. Modulation includes using amplitude modulation and frequency modulation to modulate the amplified signal to conform to a preset transmission standard and format.

[0083] The feature extraction unit extracts feature parameters from the preprocessed signal to form a signal feature vector X, including Fourier transform extraction, time-frequency analysis extraction, and wavelet transform. The synchronously extracted acquisition data groups are integrated into the signal feature vector X, which includes operating power consumption, operating time, resource usage, bit error rate, signal strength, and signal frequency.

[0084] Example 3

[0085] This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically: the communication processing module includes a data frame unit;

[0086] The data frame unit extracts data frame information from the signal feature vector X, including the data frame, packet loss rate, data reception and transmission timestamps, and total number of data packets, and calculates the packet loss rate P. loss Data latency D delay and bit error rate P error ;

[0087] The packet loss rate P loss Obtained through the following calculation formula:

[0088]

[0089] In the formula, N lost The amount of data packets lost is represented by N, specifically calculated from the difference between the number of data packets recorded by the sender and the number reported by the receiver. sent Indicates the amount of data in the data packet sent;

[0090] The data latency D delay Obtained through the following calculation formula:

[0091]

[0092] In the formula, T total N represents the total delay time, specifically obtained by summing the transmission delay times of all received data packets recorded by the receiving end. received Indicates the number of data packets received;

[0093] The bit error rate P error Obtained through the following calculation formula:

[0094]

[0095] In the formula, N error This indicates the number of bits that were erroneous, specifically calculated by comparing the received data frame with a preset transmitted data frame; N totalThis represents the total number of bits, which is specifically obtained by recording and counting the total number of bits sent at the sending end.

[0096] The system processing module includes a modeling unit;

[0097] The modeling unit uses deep learning technology to establish a control operation resource assessment model, and constructs an input layer, a hidden layer, and an output layer. The input layer receives the feature vector X and inputs the feature vector X into the control operation resource assessment model. The hidden layer performs nonlinear transformations and trains and analyzes the control operation resource assessment model. The output layer obtains the operation control energy efficiency index E. total The model parameters are iteratively adjusted using the loss function L(θ) and gradient descent algorithm.

[0098] The input layer receives feature vector X, normalizes it, and extracts the total dimension d of feature vector X to obtain the received data from the input layer: normalized feature vector X. d ={X1,X2,X3,,,,,X d};

[0099] The hidden layer consists of several hidden layers, and the output is obtained through the following calculation steps:

[0100] First hidden layer output:

[0101] H1=σ*(W1*X d +b1);

[0102] Second hidden layer output:

[0103] H2 = σ*(W2*H1+b2);

[0104] Output of the p-th hidden layer:

[0105] H p =σ*(W p *H p-1 +b p );

[0106] In the formula, W1, W2 and W p The weight matrices represent the weight matrices of the first hidden layer, the second hidden layer, and the p-th hidden layer, respectively, with b1, b2, and b... p The bias vectors represent the bias vectors of the first hidden layer, the second hidden layer, and so on up to the p-th hidden layer. σ represents the activation function, including the ReLU function, H1, H2, and H... p These represent the outputs of the first hidden layer, the second hidden layer, and so on up to the p-th hidden layer, with the p-th hidden layer output H respectively. p Specifically, the energy efficiency index E of the output layer's operation control. total .

[0107] The loss function L(θ) is obtained through the following formula:

[0108]

[0109] In the formula, L(θ) represents the loss function, specifically used to measure the preset operating control energy efficiency index and the operating control energy efficiency index E of the i-th sample. total,i The performance gap, E total,i This represents the operating control energy efficiency index of the i-th sample. This represents the preset operating control energy efficiency index, and N represents the total number of samples.

[0110] The gradient descent algorithm is adjusted using the following calculation formula:

[0111]

[0112] In the formula, θ represents the set of training parameters for controlling the operation of the resource assessment model, and α represents the learning rate, which is used to control the step size for each parameter update. This represents the gradient of the loss function with respect to the control and runtime resource evaluation model, including the partial derivatives of each parameter;

[0113] The training parameter set θ includes the weight matrix W1 (first hidden), the weight matrix W2 (second hidden), and W... p The weight matrix of the p-th hidden layer, b1 the bias vector of the first hidden layer, b2 the bias vector of the second hidden layer, and b p The bias vector of the p-th hidden layer.

[0114] In this embodiment, the input layer receives the feature vector X and inputs it into the control operation resource assessment model. The hidden layer performs nonlinear transformation and trains and analyzes the control operation resource assessment model. The output layer obtains the operation control energy efficiency index E. total The model parameters are iteratively adjusted using the loss function L(θ) and gradient descent algorithm. By utilizing layer-by-layer nonlinear transformation and iterative optimization algorithm, the energy efficiency index of the operation control is accurately calculated. The application of gradient descent algorithm ensures the precise adjustment of model parameters, thereby continuously improving the energy efficiency and performance of the system.

[0115] Example 4

[0116] This embodiment is an explanation based on Embodiment 3. Please refer to it. Figure 1 Specifically: the monitoring feedback module includes a matching unit and a control unit;

[0117] The matching unit matches preset relevant information with the required comparison value to determine the operating control energy efficiency index E. total With bit error rate Perror The system performs fitting to obtain the operating status evaluation coefficient Zxs, and matches it with the preset operating status evaluation threshold P to obtain the operating status performance evaluation scheme.

[0118] The operating status evaluation coefficient Zxs is obtained through the following calculation formula:

[0119]

[0120] In the formula, z1 and z2 represent the operating control energy efficiency index E, respectively. total With bit error rate P error The weighting coefficients, where U represents the correction constant;

[0121] The control unit performs specific execution and control according to the content of the operation status performance evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after control.

[0122] The operational status performance evaluation scheme is obtained through the following matching method:

[0123] If the operational status evaluation coefficient Zxs≤P, a qualified operational status performance evaluation result is obtained.

[0124] If the operating status evaluation coefficient Zxs > P, and the operating status performance evaluation result is unqualified, the operating parameters are adjusted, including power management parameters, data transmission parameters, cache and storage parameters, task scheduling parameters, and temperature management parameters.

[0125] Power management parameters include operating voltage, operating frequency, and power mode; data transmission parameters include transmission bandwidth, packet size, and error control; cache and storage parameters include cache size, cache policy, and storage access time; task scheduling parameters include task priority, number of threads, and scheduling policy; and temperature management parameters include heat dissipation policy and temperature threshold.

[0126] In this implementation, the energy efficiency index E is controlled through operation. total With bit error rate P error Fitting is performed to obtain the operating state evaluation coefficient Zxs, which is then matched with the preset operating state evaluation threshold P to obtain the operating state performance evaluation scheme. This enables efficient monitoring and optimization of the chip's operating state. The evaluation mechanism ensures that the system can promptly identify deficiencies in operation and formulate specific optimization measures based on the evaluation results. This helps to optimize the chip's energy efficiency and processing capabilities, and can also effectively reduce the bit error rate and improve system stability. Thus, it realizes intelligent management and dynamic optimization of the chip control system. Through this precise feedback and adjustment mechanism, the system can continuously improve performance and solve the shortcomings of traditional systems in real-time control and optimization.

[0127] Example 5

[0128] A chip-based method for optimizing dynamic signal control in IoT devices; please refer to [reference needed]. Figure 2 Specifically, it includes the following steps:

[0129] Step 1: The data acquisition module collects raw transmission signals in real time through sensors, and forms a data set by recording the power consumption, resource usage information and timestamp information during operation;

[0130] Step 2: The signal processing module preprocesses the acquired data set, including filtering, signal amplification and modulation, and simultaneously integrates the processed acquired data set into a signal feature vector X;

[0131] Step 3: The communication processing module extracts features from the signal feature vector X, including data frames, packet loss rate, and transmission / reception time information, and calculates the bit error rate P. error ;

[0132] Step 4: The system processing module uses deep learning technology to establish a control operation resource assessment model, and obtains the operation control energy efficiency index E by training and analyzing the control operation resource assessment model. total ;

[0133] Step 5: The monitoring and feedback module will control the energy efficiency index E. total With bit error rate P error The system performs fitting to obtain the operation status evaluation coefficient Zxs, and matches it with the preset operation status evaluation threshold P to obtain the operation status efficiency evaluation scheme. Then, it performs specific execution and regulation according to the content of the operation status efficiency evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after regulation.

[0134] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A chip-based dynamic signal control optimization system for Internet of Things (IoT) devices, characterized in that: It includes a data acquisition module, a signal processing module, a communication processing module, a system processing module, and a monitoring and feedback module; The data acquisition module collects raw transmission signals in real time through sensors, and forms a data set by recording the power consumption and resource information used during operation and timestamp information. The signal processing module preprocesses the acquired data set, including filtering, signal amplification, and modulation. Filtering involves using low-pass, high-pass, and band-pass filters to filter the original signal in the acquired data set, removing noise and interference. Signal amplification includes operational amplifiers and low-noise amplifiers to amplify the filtered signal to adjust its intensity. Modulation involves using amplitude modulation and frequency modulation to modulate the amplified signal, ensuring it conforms to a preset transmission standard and format. Simultaneously, the processed acquired data set is integrated into a signal feature vector X. The communication processing module extracts features from the signal feature vector X, including data frames, packet loss rate, and transmission / reception time information, and calculates the bit error rate Perror. The system processing module uses deep learning technology to establish a control operation resource evaluation model and constructs an input layer, a hidden layer, and an output layer. The input layer receives the feature vector X and inputs the feature vector X into the control operation resource evaluation model. The hidden layer performs nonlinear transformation and trains and analyzes the control operation resource evaluation model. The output layer obtains the operation control energy efficiency index Etotal and iteratively adjusts the model parameters through the loss function L(θ) and gradient descent algorithm. The monitoring and feedback module fits the operation control energy efficiency index Etotal with the bit error rate Perror to obtain the operation status evaluation coefficient Zxs, and matches it with the preset operation status evaluation threshold P to obtain the operation status efficiency evaluation scheme. Then, it performs specific execution and regulation according to the content of the operation status efficiency evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after regulation.

2. The chip-based IoT device dynamic signal control optimization system according to claim 1, characterized in that: The data acquisition module includes a signal recording unit and a resource monitoring unit; The signal recording unit captures the transmitted signals in the system in real time by using sensors and signal sensing devices, synchronously receives and records signal data from the transmission channel, and then stores the captured raw signal data in time sequence to form a raw signal data group, including component temperature, acquisition timestamp, current value, voltage value, signal strength and signal frequency. The sensors include current sensors, voltage sensors, temperature sensors, and signal strength sensors; The resource monitoring unit monitors the power consumption in real time, including CPU utilization and memory usage, synchronously records timestamp information to mark the data acquisition time, assembles power consumption data, and combines it with the original signal data group to obtain the acquired data group.

3. The chip-based IoT device dynamic signal control optimization system according to claim 1, characterized in that: The communication processing module includes a data frame unit; The data frame unit extracts data frame information from the signal feature vector X, including data frame, packet loss rate, data reception and transmission timestamps, and total number of data packets, and calculates and obtains packet loss rate Ploss, data delay rate Ddelay, and bit error rate Perror. The packet loss rate Obtained through the following calculation formula: ； In the formula, This represents the amount of data packets lost, specifically calculated by the difference between the number of data packets recorded by the sender and the number reported by the receiver. Indicates the amount of data in the data packet sent; The data latency Obtained through the following calculation formula: ； In the formula, This represents the total delay time, which is specifically obtained by summing the transmission delay times of all received data packets recorded by the receiving end. Indicates the number of data packets received; The bit error rate Obtained through the following calculation formula: ； In the formula, This indicates the number of erroneous bits, specifically calculated by comparing the received data frame with the preset transmitted data frame. This represents the total number of bits, which is specifically obtained by recording and counting the total number of bits sent at the sending end.

4. The chip-based IoT device dynamic signal control optimization system according to claim 3, characterized in that: The input layer receives feature vector X, normalizes it, and extracts the total dimension d of feature vector X to obtain the received data from the input layer: normalized feature vector. ={ , , ,、、、, }; The hidden layer consists of several hidden layers, and the output is obtained through the following calculation steps: First hidden layer output: ； Second hidden layer output: ； Output of the p-th hidden layer: ； In the formula, , and The weight matrices represent the weight matrices of the first hidden layer, the second hidden layer, and the p-th hidden layer, respectively. , and Let represent the bias vectors, specifically the bias vectors of the first hidden layer, the second hidden layer, and so on, up to the p-th hidden layer. This refers to activation functions, including the ReLU function. , and These represent the outputs of the first hidden layer, the second hidden layer, up to the p-th hidden layer, and the p-th hidden layer, respectively. Specifically, the energy efficiency index of the output layer's operation control. .

5. The chip-based IoT device dynamic signal control optimization system according to claim 1, characterized in that: The loss function Obtained through the following calculation formula: ； In the formula, This represents the loss function, specifically used to measure the preset operating control energy efficiency index and the operating control energy efficiency index of the i-th sample. The gap in performance, This represents the operating control energy efficiency index of the i-th sample. This represents the preset operating control energy efficiency index, and N represents the total number of samples. The gradient descent algorithm is adjusted using the following calculation formula: ； In the formula, This represents the set of training parameters for the resource assessment model for controlling operation. This represents the learning rate, which is used to control the step size for each parameter update. This represents the gradient of the loss function with respect to the control and runtime resource evaluation model, including the partial derivatives of each parameter; Among them, the training parameter set include The first hidden weight matrix, The second hidden weight matrix, The weight matrix of the p-th hidden layer First hidden layer bias vector, The second hidden layer bias vector and The bias vector of the p-th hidden layer.

6. The chip-based IoT device dynamic signal control optimization system according to claim 1, characterized in that: The monitoring feedback module includes a matching unit and a control unit; The matching unit matches preset relevant information with the required comparison value to determine the operating control energy efficiency index. With bit error rate The system performs fitting to obtain the operating status evaluation coefficient Zxs, and matches it with the preset operating status evaluation threshold P to obtain the operating status performance evaluation scheme. The operating status evaluation coefficient Zxs is obtained through the following calculation formula: ； In the formula, and These represent the energy efficiency index of operation control. With bit error rate The weighting coefficients, where U represents the correction constant; The control unit performs specific execution and control according to the content of the operation status performance evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after control.

7. The chip-based IoT device dynamic signal control optimization system according to claim 6, characterized in that: The operational status performance evaluation scheme is obtained through the following matching method: If the operational status evaluation coefficient Zxs≤P, a qualified operational status performance evaluation result is obtained. If the operating status evaluation coefficient Zxs > P, and the operating status performance evaluation result is unqualified, the operating parameters, including power management parameters, data transmission parameters, cache and storage parameters, task scheduling parameters, and temperature management parameters, will be adjusted.

8. A chip-based dynamic signal control optimization method for IoT devices, comprising the chip-based dynamic signal control optimization system for IoT devices as described in any one of claims 1-7, characterized in that: Includes the following steps: Step 1: The data acquisition module collects raw transmission signals in real time through sensors, and forms a data set by recording the power consumption, resource usage information and timestamp information during operation; Step 2: The signal processing module preprocesses the acquired data set, including filtering, signal amplification and modulation, and simultaneously integrates the processed acquired data set into a signal feature vector X; Step 3: The communication processing module extracts features from the signal feature vector X, including data frames, packet loss rate, and transmission / reception time information, and calculates the bit error rate. ; Step 4: The system processing module uses deep learning technology to establish a control operation resource assessment model, and obtains the operation control energy efficiency index by training and analyzing the control operation resource assessment model. ; Step 5: The monitoring and feedback module will run the control energy efficiency index. With bit error rate The system performs fitting to obtain the operation status evaluation coefficient Zxs, and matches it with the preset operation status evaluation threshold P to obtain the operation status efficiency evaluation scheme. Then, it performs specific execution and regulation according to the content of the operation status efficiency evaluation scheme, and simultaneously monitors and iteratively evaluates the operation control information after regulation.

Images