Key generation method and system based on motion state adaptive Kalman filtering
Through adaptive Kalman filtering technology, the Kalman filter parameters of the drone are adjusted in real time during flight, which solves the problem of decreased accuracy of traditional Kalman filtering in high dynamic environments and improves the key generation efficiency and security in drone communications.
Patent Information
- Application Number
- CN202510822568.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-19
- Publication Date
- 2025-09-19
AI Technical Summary
The fixed parameters and environmental sensitivity of traditional Kalman filtering in highly dynamic environments such as drones lead to reduced estimation accuracy, making it difficult to meet the needs of highly dynamic scenes.
An adaptive Kalman filtering method based on motion state is adopted. The dynamic characteristic indicators are calculated by real-time perception of flight state data, and predefined nonlinear functions are used to map them into Kalman filtering parameters. The covariance matrix of process noise and observation noise is dynamically adjusted. The adaptive Kalman filtering is performed in combination with channel state information to generate a key.
The channel prediction accuracy and reciprocity are improved, and the key generation efficiency and security are significantly improved.
Smart Images
Figure CN120676349A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wireless communication security technology, and in particular to a key generation and identification method and system in wireless communication, which is particularly suitable for unmanned aerial vehicle (UAV) mobile communication scenarios. Background Art
[0002] With the widespread adoption of wireless communication technologies in areas such as drones, vehicle-to-everything (V2X), smart cities, and the Industrial Internet of Things (IIoT), communication security has become a crucial research topic. In these mobile environments, wireless channels experience significant dynamic variations, including multipath fading, obstruction, and interference. These factors pose significant challenges to traditional upper-layer encryption mechanisms. Furthermore, drone and IoV communications often involve highly dynamic environments, such as high-speed vehicle movement or aerial drone flight. This significantly reduces the security and applicability of traditional static key generation and distribution mechanisms.
[0003] As a classic state estimation algorithm, Kalman filtering is widely used in fields such as aircraft navigation and target tracking. However, in complex flight environments, traditional Kalman filtering has the following problems:
[0004] 1. Parameter fixation: Filter parameters (such as the process noise covariance matrix Q and the observation noise covariance matrix R) are difficult to adapt to the dynamic changes of flight conditions, resulting in a decrease in estimation accuracy;
[0005] 2. Environmental sensitivity: Traditional filtering algorithms are prone to divergence in scenarios such as high-speed maneuvers and strong disturbances.
[0006] Existing solutions often rely on empirical parameter adjustments or fixed rules, lacking real-time responsiveness to flight conditions and struggling to meet the demands of highly dynamic scenarios. Therefore, an adaptive parameter modulation method is urgently needed to improve filter reciprocity and enhance algorithm robustness. Summary of the Invention
[0007] Purpose of the invention: To solve the above problems, the present invention proposes a key generation method and system based on motion state adaptive Kalman filtering, which improves CSI reciprocity through adaptive Kalman filtering technology and dynamically adjusts filtering parameters based on flight status to improve key generation efficiency and security.
[0008] Technical solution: In order to achieve the above invention objectives, the present invention adopts the following technical solution:
[0009] In a first aspect, a method for generating a key based on a motion state adaptive Kalman filter comprises the following steps:
[0010] A method for generating a key based on a motion state adaptive Kalman filter, characterized by comprising the following steps:
[0011] Calculating dynamic characteristic indicators based on real-time sensed flight status data, wherein the flight status data includes speed, acceleration, and attitude angle, and the dynamic characteristic indicators include speed change rate and angular velocity;
[0012] Based on the pre-trained weight coefficients, the dynamic feature indicators are mapped to Kalman filter parameters using a predefined nonlinear function, including the process noise covariance matrix Q adj , observation noise covariance matrix R adj ;
[0013] The channel state information CSI of the wireless channel is used as the state data to perform adaptive Kalman filtering. In the prediction stage, the parameter Q is dynamically adjusted according to the flight state. adj , update the state prediction X k|k-1 and the prior covariance P k|k-1 ; In the update phase, combined with the channel state information observation value Z at time k k , using R adj Calculate the Kalman gain K k , update the state estimate X k and covariance P k ;
[0014] The filtered channel state information CSI is subjected to energy normalization, median quantization, Gray code conversion, LDPC encoding and hash privacy amplification to generate the final key.
[0015] Furthermore, the velocity change rate is expressed as:
[0016]
[0017] The rate of change of angular velocity is expressed as:
[0018]
[0019] Among them, Δt is the time period, Δv is the velocity change, θ is the pitch angle, Δω θ is the change in angular velocity in the pitch angle direction
[0020] Furthermore, a predefined nonlinear function is used to map the dynamic characteristic index to the Kalman filter parameter, which is expressed as follows:
[0021] Q adj =Q0·(1+k1·|v|+k2·α v )
[0022] r adj =R0·(1+k3·|φ|+k4·β θ )
[0023] Among them, k1~k4 are pre-training weight coefficients; α vis the velocity change rate, β θ is the angular velocity, v is the velocity, φ is the roll angle, Q0 and R0 are the preset basic values.
[0024] Furthermore, the weight coefficients k1 to k4 are pre-trained in the following way:
[0025] Construct a flight status dataset with the input being [||v||,α v ,|φ|,|θ|]; v is the acceleration, α v is the velocity change rate, θ is the pitch angle, and φ is the roll angle;
[0026] Train a multi-output deep neural network DNN, and use the Sigmoid function to constrain the four weight coefficients of the output layer to the range of (0,1);
[0027] The weights are adjusted with the optimization goal of minimizing the reciprocity error between the uplink channel state information and the downlink channel state information.
[0028] Furthermore, the adaptive Kalman filter satisfies:
[0029] Prediction stage:
[0030] Update phase:
[0031] Among them, F k is the state transition matrix, which describes how the system transfers from time k-1 to time k; P k-1 is the covariance at time k-1; K k is the Kalman gain; H k is the state observation matrix, which describes the mapping relationship between the state space and the observation space; Z k is the actual CSI observation value at time k, H k X k|k-1 is the predicted observed value.
[0032] Furthermore, key generation specifically includes:
[0033] The channel state information (CSI) amplitude value is normalized to the range [0, 1]. Binary quantization is performed with the median as the threshold. The quantized bits are converted into Gray code. Information reconciliation is achieved through LDPC coding. The SM3 hash function is used for privacy amplification.
[0034] Furthermore, the channel state information CSI is exchanged in a time slice rotation manner.
[0035] In a second aspect, a key generation system based on motion state adaptive Kalman filtering comprises:
[0036] A feature extraction module is used to calculate dynamic feature indicators based on real-time sensed flight status data, wherein the flight status data includes speed, acceleration, and attitude angle, and the dynamic feature indicators include speed change rate and angular velocity;
[0037] Parameter mapping module, which is used to map dynamic feature indicators into Kalman filter parameters based on pre-trained weight coefficients using predefined nonlinear functions, including the process noise covariance matrix Q adj , observation noise covariance matrix R adj ;
[0038] The adaptive Kalman filter module is used to perform adaptive Kalman filtering using the channel state information (CSI) of the wireless channel as state data. In the prediction stage, the parameter Q is dynamically adjusted according to the flight status. adj , update the state prediction X k|k-1 and the prior covariance P k|k-1 ; In the update phase, combined with the channel state information observation value Z at time k k , using R adj Calculate the Kalman gain K k , update the state estimate X k and covariance P k ;
[0039] The key generation module is used to perform energy normalization, median quantization, Gray code conversion, LDPC encoding and hash privacy amplification on the filtered channel state information CSI to generate the final key.
[0040] In a third aspect, an electronic device comprises: a memory; one or more processors; and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and when the programs are executed by the processors, the key generation method based on motion state adaptive Kalman filtering as described in the first aspect is implemented.
[0041] In a fourth aspect, a computer-readable storage medium stores a computer program thereon, wherein when the computer program is executed by a processor, the method for generating a key based on motion state adaptive Kalman filtering as described in the first aspect is implemented.
[0042] The beneficial effects of the present invention are: by sensing the aircraft's motion state (speed, acceleration, attitude angle) in real time, mapping it into Kalman filter parameters, and combining the channel state information to perform an adaptive Kalman filter process, the Kalman filter parameters are dynamically adjusted, the channel prediction accuracy and reciprocity are improved, and the key generation efficiency and security are significantly improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1This is a hardware architecture diagram of the sensor and CSI acquisition system of the present invention;
[0044] Figure 2 This is a schematic diagram of the first process of system initialization and state perception;
[0045] Figure 3 This is a CSI acquisition timing diagram based on ESP32;
[0046] Figure 4 This is a schematic diagram of the second process of Kalman filter state update;
[0047] Figure 5 This is a schematic diagram of the third process of key generation and secure communication;
[0048] Figure 6 It is a flowchart of the key generation method of the present invention. DETAILED DESCRIPTION
[0049] In order to have a clearer understanding of the features and advantages of the technical solution of the present invention, the composition and implementation of the specific solution are explained below with reference to the accompanying drawings.
[0050] The embodiment of the present invention provides a channel reciprocity improvement method based on motion state adaptive Kalman filtering, which is applied to two drones equipped with ESP32 wireless communication modules (hereinafter referred to as drone A and drone B). Figure 1 As shown, in the system architecture of this embodiment, each drone includes the following hardware configuration:
[0051] Communication main control unit: uses ESP32-WROOM-32U microcontroller to complete the main calculation and control processes;
[0052] Motion sensing unit: using MPU6050 six-axis sensor;
[0053] Positioning unit: using high-precision GPS module;
[0054] Wireless communication unit: ESP32, built-in 802.11n WiFi module (external high-gain antenna);
[0055] Power amplifier unit: uses a USB power supply interface and is connected to the ESP32 signal output port to amplify the signal power.
[0056] The first process: system initialization and status perception
[0057] As attached Figure 2 As shown, in the first process of this embodiment, the following steps are included:
[0058] Step 11: System Initialization
[0059] Drone A and Drone B are powered on and the FreeRTOS real-time operating system is loaded on ESP32.
[0060] Create a dual-task thread:
[0061] High-priority tasks (5ms cycle): sensor data acquisition and Kalman filtering.
[0062] Low-priority tasks (10ms period): key generation and encrypted communication.
[0063] Initialize the Kalman filter parameters:
[0064] State vector: X0 = [0, 0, 0] T ;
[0065] Error covariance: P0 = diag(1,1,1);
[0066] Basic noise parameters: Q0 = 0.001, R0 = 0.01;
[0067] Step 12: Flight Status Perception
[0068] Real-time collection of motion data via MPU6050:
[0069] Three-dimensional acceleration: a=[a x ,a y ,a z ];
[0070] Attitude angles (pitch angle θ, roll angle φ);
[0071] Calculate dynamic characteristic indicators:
[0072] The calculated speed change rate is expressed as:
[0073]
[0074] The calculated rate of change of angular velocity is expressed as:
[0075]
[0076] where Δω θ is the change in angular velocity in the pitch angle direction.
[0077] Step 3: CSI collection and preprocessing
[0078] Enable the CSI acquisition function of ESP32 and use time slice rotation to interact with CSI.
[0079] The key generation technology based on wireless channel characteristics requires both communicating parties to send and receive signals in a short time interval at the shortest possible time interval to obtain the wireless channel characteristics. This example uses the most common basic access mode {Data, ACK} to perform channel detection between the communicating parties. Through uplink and downlink interaction frames with microsecond-level deterministic intervals, the communicating parties complete the measurement of the highly reciprocal wireless channel. The specific interaction process is as follows: Figure 3 As shown in the figure, the AP first sends reconciliation information to the STA via UDP (if it is the first time, it will be none). After receiving the reconciliation information, the STA processes it and obtains a reconciliation result (yes or no). The reconciliation result is then sent to the AP via UDP. The AP receives the transmitted DATA frame carrying the reconciliation result and obtains the CSI. The AP also returns an ACK frame to the STA. After receiving the ACK frame, the STA records the current CSI.
[0080] Second process: Adaptive Kalman filtering
[0081] As attached Figure 4 As shown, in the second process of this embodiment, the following steps are included:
[0082] Step 21: Design an offline weight coefficient pre-training scheme based on deep learning
[0083] S21-1. Constructing a flight status dataset
[0084] Multi-dimensional flight status data is collected through the MPU6050 and GPS sensors. This data is then fed into the data preprocessing module for necessary pre-processing, such as normalization, scaling, and outlier handling, to ensure that the data format and range meet the model input requirements.
[0085] S21-2. Training the neural network mapping model
[0086] The model is a multi-output deep neural network (DNN).
[0087] Input layer: receives the preprocessed 4-dimensional flight state vector [||v||,α v ,|φ|,|θ|].
[0088] Hidden layers: These layers contain multiple fully connected layers that use activation functions (ReLU) and nonlinear transformations to extract deep features from the input state. Batch normalization layers are used to stabilize training and accelerate convergence.
[0089] Output layer: Designed as four independent neurons, each corresponding to a weight coefficient (k1, k2, k3, k4). A Sigmoid activation function is used to constrain the output range of each coefficient to be between (0, 1), ensuring that the adjustment of the basic noise parameters has reasonable limits.
[0090] Step 22: Dynamic Mapping of Parameters
[0091] Dynamically adjust the filter parameters according to the flight status by [||v||,α v ,|φ|,|θ|] is used as input to look up the table to obtain k1~k4, and the adjusted process noise covariance matrix Q under the current state is dynamically calculated based on the coefficients and state information. adj and the observation noise covariance matrix R adj .
[0092] Q adj =Q0·(1+k1·|v|+k2·α v )
[0093] R adj =R0·(1+k3·|φ|+k4·β θ )
[0094] Among them, Q0 and R0 are the preset basic values.
[0095] Step 23: State Prediction
[0096] Since the interval between two adjacent CSI frames is about 10ms, in order to simplify the model, the predicted state X of the kth frame CSI is k|k-1 Think with X k-1 Same. Update state predictions and covariances based on the system model:
[0097] X k|k-1 =F k X k-1 +B k u k
[0098]
[0099] Among them, F k is the state transition matrix, which describes how the system transfers from time k-1 to time k; B k The control matrix describes how the input affects the system state; u k is the control input at time k. In this case, as a pure perception system, u k Normally 0 (no control input). X k-1 Refers to the CSI at time k-1.
[0100] Step 24: Kalman filter state update
[0101] Update phase: Combined with the channel state information observation value Z k , calculate the Kalman gain K k :
[0102] K k =P k|k-1 ·(P k|k-1 +R adj ) -1
[0103] Update state estimates and covariances:
[0104] X k =X k|k-1 +K k (Z k -X k|k-1 )
[0105] P k =(IK k )P k|k-1
[0106] Among them, K k is the Kalman gain, H k is the state-observation matrix, which describes how the system maps the state space to the observation space, Z k is the actual CSI observation value at time k, H k X k|k-1 is the predicted observed value.
[0107] S25. Reciprocity error calculation:
[0108] Calculate the difference (mean square error (MSE)) between the filtered uplink and downlink CSI estimates. This difference is the reciprocity error (RE)—a key indicator for measuring filter performance. The smaller this error, the more consistent the uplink and downlink channel estimates are, and the better the reciprocity.
[0109] The third process: key generation and secure communication
[0110] As attached Figure 5 As shown, in the third process of this embodiment, the following steps are included:
[0111] Step 31: Perform energy normalization on the CSI obtained in the second process;
[0112] Step 31: Use median quantization to quantize the CSI to obtain a quantized value;
[0113] Step 31: Convert the quantized value into Gray code to reduce bit jumps.
[0114] Step 31: The obtained Gray code is LDPC (Low-Density Parity-Check Code) encoded, and UAV A and UAV B reconcile information through a common channel.
[0115] Furthermore, the SM3 hash function is applied to amplify the privacy of the key.
[0116] Repeat the steps from the first process to the third process above, and continuously correct the Kalman filter gain until the task is completed.
[0117] According to the system architecture and processing described in detail above, it can be understood that for each drone's communication main control unit, it is configured to execute a key generation method based on motion state adaptive Kalman filtering, referring to Figure 6 , the method comprises the following steps:
[0118] Calculating dynamic characteristic indicators based on real-time sensed flight status data, wherein the flight status data includes speed, acceleration, and attitude angle, and the dynamic characteristic indicators include speed change rate and angular velocity;
[0119] Based on the pre-trained weight coefficients, the dynamic feature indicators are mapped to Kalman filter parameters using a predefined nonlinear function, including the process noise covariance matrix Q adj , observation noise covariance matrix R adj ;
[0120] The channel state information CSI of the wireless channel is used as the state data to perform adaptive Kalman filtering. In the prediction stage, the parameter Q is dynamically adjusted according to the flight state. adj , update the state prediction X k|k-1 and the prior covariance P k|k-1 ; In the update phase, combined with the channel state information observation value Z at time k k , using R adj Calculate the Kalman gain K k , update the state estimate X k and covariance P k ;
[0121] The filtered channel state information CSI is subjected to energy normalization, median quantization, Gray code conversion, LDPC encoding and hash privacy amplification to generate the final key.
[0122] The specific calculation process in each step will not be repeated here.
[0123] Another embodiment of the present invention further provides a key generation system based on motion state adaptive Kalman filtering, comprising:
[0124] A feature extraction module is used to calculate dynamic feature indicators based on real-time sensed flight status data, wherein the flight status data includes speed, acceleration, and attitude angle, and the dynamic feature indicators include speed change rate and angular velocity;
[0125] Parameter mapping module, which is used to map dynamic feature indicators into Kalman filter parameters based on pre-trained weight coefficients using predefined nonlinear functions, including the process noise covariance matrix Q adj , observation noise covariance matrix R adj ;
[0126] The adaptive Kalman filter module is used to perform adaptive Kalman filtering using the channel state information (CSI) of the wireless channel as state data. In the prediction stage, the parameter Q is dynamically adjusted according to the flight status. adj , update the state prediction X k|k-1 and the prior covariance P k|k-1 ; In the update phase, combined with the channel state information observation value Z at time k k , using R adj Calculate the Kalman gain K k , update the state estimate X k and covariance P k ;
[0127] The key generation module is used to perform energy normalization, median quantization, Gray code conversion, LDPC encoding and hash privacy amplification on the filtered channel state information CSI to generate the final key.
[0128] The specific calculation process in each module will not be repeated here.
[0129] Another embodiment of the present invention also provides an electronic device, comprising: a memory; one or more processors; and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and when the programs are executed by the processors, the key generation method based on motion state adaptive Kalman filtering as described above is implemented.
[0130] Another embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the key generation method based on motion state adaptive Kalman filtering as described above is implemented.
[0131] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, devices (systems), electronic devices, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0132] The present invention is described with reference to flowcharts of methods according to embodiments of the present invention. It should be understood that each process in the flowcharts and combinations of processes in the flowcharts can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts. Figure 1 A device that specifies functions in a process or multiple processes.
[0133] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A function specified in a process or multiple processes.
[0134] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 The steps of a specified function in a process or multiple processes.
Claims
1. A key generation method based on motion state adaptive Kalman filtering, characterized in that: The following steps are involved: Calculating dynamic characteristic indicators based on real-time sensed flight status data, wherein the flight status data includes speed, acceleration, and attitude angle, and the dynamic characteristic indicators include speed change rate and angular velocity; Based on the pre-trained weight coefficients, the dynamic feature indicators are mapped to Kalman filter parameters using a predefined nonlinear function, including the process noise covariance matrix Q adj , observation noise covariance matrix R adj ; Using the channel state information CSI of the wireless channel as the state data, an adaptive Kalman filter is performed. In the prediction stage, the parameter Q is dynamically adjusted according to the flight state. adj , update the state prediction X k|k-1 and the prior covariance P k|k-1 ; In the update phase, combined with the channel state information observation value Z at time k k , using R adj Calculate the Kalman gain K k , update the state estimate X k and covariance P k ; The filtered channel state information CSI is subjected to energy normalization, median quantization, Gray code conversion, LDPC encoding and hash privacy amplification to generate the final key.
2. The method according to claim 1, characterized in that The rate of change of velocity is expressed as: The rate of change of angular velocity is expressed as: Among them, Δt is the time period, Δv is the velocity change, θ is the pitch angle, Δω θ is the change in angular velocity in the pitch angle direction.
3. The method according to claim 1, characterized in that The dynamic characteristic index is mapped to the Kalman filter parameter using a predefined nonlinear function, and the expression is: Q adj =Q0·(1+k1·|v|+k2·α v ) R adj =R0·(1+k3·|φ|+k4·β θ ) Among them, k1~k4 are pre-training weight coefficients; α v is the velocity change rate, β θ is the angular velocity, v is the velocity, φ is the roll angle, Q0 and R0 are the preset basic values.
4. The method according to claim 1, wherein The weight coefficients k1 to k4 are pre-trained in the following way: Construct a flight status dataset with the input being [||v||,α v ,|φ|,|θ|]; v is the velocity, α v is the velocity change rate, θ is the pitch angle, and φ is the roll angle; Train a multi-output deep neural network DNN, and use the Sigmoid function to constrain the four weight coefficients of the output layer to the range of (0,1); The weights are adjusted with the optimization goal of minimizing the reciprocity error between the uplink channel state information and the downlink channel state information.
5. The method according to claim 1, wherein Adaptive Kalman filtering satisfies: Prediction stage: Update phase: Among them, F k is the state transition matrix, which describes how the system transfers from time k-1 to time k; P k-1 is the covariance at time k-1; K k is the Kalman gain; H k is the state observation matrix, which describes the mapping relationship between the state space and the observation space; Z k is the actual CSI observation value at time k, H k X k|k-1 is the predicted observed value.
6. The method according to claim 1, characterized in that Key generation specifically includes: The channel state information (CSI) amplitude value is normalized to the range [0, 1]. Binary quantization is performed with the median as the threshold. The quantized bits are converted into Gray code. Information reconciliation is achieved through LDPC coding. The SM3 hash function is used for privacy amplification.
7. The method according to claim 1, characterized in that The channel state information (CSI) is exchanged in a time slice rotation manner.
8. A key generation system based on motion state adaptive Kalman filtering, characterized in that: include: A feature extraction module is used to calculate dynamic feature indicators based on real-time sensed flight status data, wherein the flight status data includes speed, acceleration, and attitude angle, and the dynamic feature indicators include speed change rate and angular velocity; Parameter mapping module, which is used to map dynamic feature indicators into Kalman filter parameters based on pre-trained weight coefficients using predefined nonlinear functions, including the process noise covariance matrix Q adj , observation noise covariance matrix R adj ; The adaptive Kalman filter module is used to perform adaptive Kalman filtering using the channel state information (CSI) of the wireless channel as state data. In the prediction stage, the parameter Q is dynamically adjusted according to the flight status. adj , update the state prediction X k|k-1 and the prior covariance P k|k-1 ; In the update phase, combined with the channel state information observation value Z at time k k , using R adj Calculate the Kalman gain K k , update the state estimate X k and covariance P k ; The key generation module is used to perform energy normalization, median quantization, Gray code conversion, LDPC encoding and hash privacy amplification on the filtered channel state information CSI to generate the final key.
9. An electronic device, characterized in that: include: Memory; one or more processors; And one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and when the programs are executed by the processors, the key generation method based on motion state adaptive Kalman filtering as described in any one of claims 1-7 is implemented.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for generating a key based on motion state adaptive Kalman filtering according to any one of claims 1 to 7 is implemented.