A Resource Scheduling-Based Integrated Radar-Communication Method for UAV Swarms

Abstract

This invention provides a resource-scheduling-based integrated radar and communication method for UAV swarms. By allocating resources such as the working center frequency, bandwidth, and power, it optimizes the design of the integrated radar and communication system and significantly improves the covert communication and detection performance of the UAV system. The key technical steps of this invention include: first, constructing an integrated radar and communication system for the UAV swarm based on its covert communication and detection performance requirements; second, employing a resource scheduling method to process the system; and finally, constructing an optimized resource allocation method based on a combination of reinforcement learning and water-filling methods to effectively allocate system resources. Utilizing the integrated radar and communication technology of this invention, the resource utilization efficiency of UAV swarms can be effectively improved, enhancing the combined performance of detection and covert communication.

Description

Technical Field

[0001] This invention belongs to the fields of radar and signal processing technology and artificial intelligence technology, and particularly relates to an integrated radar communication method for UAV swarms based on a resource scheduling algorithm using reinforcement learning and water injection method. Background Technology

[0002] Drone swarms are now widely used in many areas of life, including agriculture, geology, military, and urban management. With the advancement of drone swarm applications, the requirements for drone swarms are also increasing. Among these requirements, the security of communication between drone swarms is paramount. This is mainly because communication between drone swarms is broadcast-based, making it easily intercepted and eavesdropped on, which could have significant consequences.

[0003] Currently, wireless communication security technologies mainly focus on two directions: encryption and covert communication. Regarding encryption, some scholars have proposed using private networks, essentially physically isolating users. However, with the development of current communication networks, a relatively public setting is required, significantly limiting the applicability of this method. Therefore, some scholars have recently proposed applications for UAV swarm communication security, primarily studying physical layer key generation and key negotiation techniques from an information theory security perspective. They have proposed secure communication schemes suitable for UAV swarm wireless networks and analyzed the communication performance of UAV swarm networks under specific models. However, this encryption method is based on the assumption that the wireless channel transmits error-free data and that eavesdroppers cannot decipher the key. Due to the broadcast nature, fading, and openness of wireless channels, as well as the presence of various noises and interference, error-free transmission is difficult. Furthermore, if the physical layer is completely transparent to eavesdroppers, coupled with the increased computational complexity of modern computers, the consequences of deciphering the key would be unimaginable. In addition, some scholars have proposed covert communication methods using images as the channel, employing a predictive difference expansion method. However, this method is susceptible to being analyzed by hacking tools, leading to significant security vulnerabilities. Of course, there are other common technical methods such as spread spectrum technology. The advantages of spread spectrum technology are strong anti-interference capability and good concealment performance, while its disadvantages include wide system bandwidth and complex system implementation.

[0004] In recent years, radar-communication integration technology has played an increasingly important role in both military and civilian fields due to its low cost, light weight, and high integration. Its essence lies in simultaneously realizing radar and communication functions on a unified and shared hardware platform. Currently, radar-communication integration is mainly achieved through three approaches: shared waveforms, time-division multiplexing, and beam splitting.

[0005] Based on the limitations of the above technologies and the development of radar-communication integration technology, this invention, from the perspective of physical layer resource allocation, equips radar-communication integration equipment into the UAV swarm system, which not only reduces weight and cost, but also improves the covert communication between UAV swarms and the detection performance of interceptors. Summary of the Invention

[0006] The purpose of this invention is to provide an integrated radar and communication system and method for UAV swarms based on resource scheduling. By allocating resources such as working center frequency, bandwidth and power, the integrated radar and communication system is optimized and the covert communication and detection performance of the UAV system is significantly improved.

[0007] The specific technical solution of the present invention is as follows:

[0008] A method for integrating radar and communication in UAV swarms based on resource scheduling includes the following steps:

[0009] Step 1: Based on the performance requirements of covert communication and detection of UAV swarms, construct an integrated radar and communication system for UAV swarms;

[0010] Step 2: Process the system using resource scheduling methods;

[0011] Step 3: In order to effectively allocate system resources, we construct an optimized resource allocation method based on a combination of reinforcement learning and water injection method.

[0012] Furthermore, the integrated radar and communication system for the UAV swarm described in step 1 is as follows:

[0013] The system primarily consists of multiple drones equipped with integrated radar and communication systems. When enemy interceptors are unable to effectively intercept the drones, these drones conduct covert communication with each other based on the detection, identification, and location of the interceptor. Furthermore, the integrated radar and communication systems allow for the efficient allocation of resources for communication and target detection among the drones.

[0014] Furthermore, the conditions under which the enemy interceptor aircraft cannot effectively intercept the target, as described in step 1, are as follows:

[0015] By optimizing the design of radar and communication resources, the enemy interceptor aircraft is unable to effectively intercept the wireless communication between the drone swarms. The enemy interceptor aircraft uses hypothesis testing to identify the intercepted signals. By optimizing resource allocation, the identification error rate is maximized. The error rate is composed of the false alarm probability and the missed alarm probability.

[0016] The final formula is as follows:

[0017]

[0018] Where B represents the number of channels; n represents the channel number; Δω kw P represents the detection error of UAV k against interceptor w; kk′ λ represents the power allocated for communication between the k-th UAV and the k′-th UAV. n This represents the wavelength corresponding to the assigned channel n; g kn Indicates whether the nth channel is assigned to the kth UAV for radar detection; β represents the channel gain coefficient; u k This represents the location information of drone k; ω w This indicates the location information of the interceptor ω; This represents the maximum resource limit under the condition of maximum interception error rate; σ 2 This represents the noise variance.

[0019] Furthermore, the covert communication performance indicators of the UAV swarm system described in step 1 are as follows:

[0020]

[0021] (Δω kw ) 2 ≤δ

[0022]

[0023] Where K is the number of drones in the swarm; k is the individual drone number in the swarm; k′ is the communication target of drone k; N represents the number of channels; n represents the channel number; P kk′ σ represents the power allocated for communication between the k-th drone and the k′-th drone; 2 h represents the noise variance. kk′n This represents the channel gain allocated to the communication between the k-th UAV and the k′-th UAV using channel n. This represents the probability that the most recently intercepted drone is true when drone k is communicating. This represents the probability that the most recently intercepted device was detected as a fake. kk′n This indicates whether the nth channel is allocated for communication between the kth and k′th drones; 0 indicates no allocation, and 1 indicates allocation. kn Indicates whether the nth channel is allocated to the kth UAV for interception detection; δ represents the radar detection threshold; Δω kw λ represents the detection error of UAV k against interceptor w; n P represents the wavelength corresponding to the assigned channel n; kk′r This represents the received power when the k-th drone communicates with the k′-th drone; L represents the square of the distance from which drone k detected the nearest interceptor; kk′G represents the system loss coefficient for communication between the k-th UAV and the k′-th UAV; kk′t and G kk′r These represent the transmit gain and receive gain, respectively. β represents the channel gain coefficient; u k This represents the location information of drone k; ω w This indicates the location information of the interceptor ω; This represents the maximum resource limit under the condition of the highest interception error rate.

[0024] Furthermore, the system resources and their allocation methods corresponding to the resource scheduling method in step 2 are as follows:

[0025] Resources mainly include the operating center frequency, bandwidth, and power. Bandwidth refers to the bandwidth at the corresponding operating center frequency. Furthermore, each bandwidth is divided into multiple channels, and each channel cannot be repeatedly allocated. Additionally, power resources limit the energy allocated to each UAV for radar detection and wireless communication. Finally, the free combination of these three resources constitutes a resource block.

[0026] System resource allocation method: Each communication and target detection among the drone clusters will be allocated a resource block.

[0027] Furthermore, the main limitations of the resource scheduling method described in step 2 are as follows:

[0028] Based on the most basic communication requirements, covert communication has a minimum communication rate limit:

[0029]

[0030] Where N represents the number of channels; n represents the channel number; P kk′ h represents the power allocated for communication between the k-th drone and the k′-th drone. kk′n The channel gain σ represents whether the nth channel is allocated to communication between the kth and k′ drones; 2 Indicates the noise figure; This indicates the threshold for the most basic communication requirements of a drone.

[0031] Based on the most basic detection requirements, radar detection has minimal detection resource constraints.

[0032]

[0033] Where N represents the number of channels; n represents the channel number; T p Indicates pulse width; v represents the radar detection RCS; P k h represents the power allocated to the k-th UAV during its reconnaissance mission. knσ represents the channel gain when the nth channel is allocated to the kth UAV for detection; 2 Indicates the noise figure; This indicates the threshold for the most basic radar detection requirements of drones.

[0034] Furthermore, step 3 describes the construction of an optimized resource allocation method based on a combination of reinforcement learning and the water-filling method. For resources such as the operating center frequency and corresponding bandwidth, reinforcement learning is used for allocation; for resources such as power, the water-filling method is used for allocation.

[0035] Constructing an environment model for reinforcement learning includes environmental state, agent behavior, and reward;

[0036] Environmental state: ε [t] It indicates the current environmental state and is composed of the communication links of the UAV swarm and the channels allocated for target detection.

[0037]

[0038] in, Represents a real dataset; K represents the number of drone swarms; M represents... [t] ={m kk′ ,k=1,2,…,K} represents the channel allocated for the communication link between drone swarms. m kk′ E represents the channel allocated for communication between drone k and drone k′. [t] ={e k ,k=1,2,…,K} represents the channel assigned to each UAV when performing target detection. k This represents the channel assigned to the UAV k when it performs target detection.

[0039] Agent behavior: in state ε [t] Let A be all possible behaviors. [t] .

[0040]

[0041] The number of actions is The f-th action is defined as This represents the o-th channel at the j-th operating frequency. [t] and O [t] Let A and B represent the allocatable working center frequency and channel in the current state, respectively. The agent selects an action from the action set based on the current state and reward. Note that, considering channel interference, action set A... [t] Actions that use the same channel as the previous actions are not included.

[0042] Reward: Based on the current resource allocation strategy ε [t]The reward is defined as a performance indicator of the covert communication of the drone swarm system, and our goal is to maximize the reward.

[0043] Constructing a model for the water injection method:

[0044] First, assuming the target detection power is a fixed value, calculate the communication power between the drone swarm:

[0045]

[0046] Among them, P kk′ f represents the power allocated for communication between drone k and drone k′. k This represents the probability of identifying a drone as real, assuming the nearest interceptor to drone k is real. kk′S This represents the number of channels allocated for communication between UAV k and UAV k′ when the operating center frequency is S. L1 represents the Lagrange multiplier corresponding to the communication power. r This indicates that the target detection power is a fixed value at this time. This represents the total power limit of drone k.

[0047] Then, based on the obtained communication power, calculate the target detection power:

[0048]

[0049] Among them, P k f represents the power allocated to UAV k for target detection. kk′ This represents the communication rate between drone k and drone k′. kS This represents the number of channels allocated to UAV k for target detection when the operating center frequency is S. L2 represents the Lagrange multiplier corresponding to the target detection power. kk′ This represents the power allocated for communication between drone k and drone k′. This represents the total power limit of drone k.

[0050] The present invention has the following advantages:

[0051] Compared to traditional encryption methods, this invention eliminates the need to consider error-free transmission over wireless channels and computational complexity, making it applicable to various integrated radar-communication applications for UAV swarms. Secondly, compared to spread spectrum and chaotic technologies, it has lower resource requirements and a simpler system design. Finally, due to its integrated radar-communication approach, overall resource utilization is significantly improved. Attached Figure Description

[0052] Figure 1 This is a diagram of the integrated radar and communication technology system for unmanned aerial vehicle (UAV) swarms according to the present invention.

[0053] Figure 2 This is a resource diagram for the present invention;

[0054] Figure 3 This is the resource scheduling diagram of the present invention;

[0055] Figure 4 This invention employs reinforcement learning to achieve convergence of operating frequency and bandwidth allocation.

[0056] Figure 5 This is a performance comparison chart of the UAV swarm radar-communication integrated technology under varying channel numbers.

[0057] Figure 6 This is a performance comparison chart of the integrated radar and communication technology for UAV swarms under varying UAV numbers, based on the present invention. Detailed Implementation

[0058] To better understand the purpose, structure, and function of this invention, the following detailed description of a resource-scheduling-based integrated radar and communication technology for unmanned aerial vehicle (UAV) swarms is provided in conjunction with the accompanying drawings.

[0059] The specific process of this invention is as follows:

[0060] 1) such as Figure 1 The diagram shows an integrated radar and communication system for a drone swarm. The system primarily consists of multiple drones equipped with integrated radar and communication equipment. These drones form a swarm of K drones. Under conditions where enemy interceptors cannot effectively intercept them, the multiple drones conduct covert communication with each other based on the detection, identification, and location of the interceptor. Furthermore, the integrated radar and communication equipment allows for the rational allocation of resources for communication and target detection among the multiple drones.

[0061] 2) Based on the system, construct the covert communication performance indicators of the UAV swarm system. The performance indicators are based on the effective identification and detection of the interceptor under the condition that the enemy interceptor cannot effectively intercept it.

[0062] The process by which enemy interceptors are unable to effectively intercept a target is as follows:

[0063] By optimizing the design of radar and communication resources, the enemy interceptor aircraft is unable to effectively intercept the wireless communication between the drone swarms. The enemy interceptor aircraft uses hypothesis testing to identify the intercepted signals. By optimizing resource allocation, the identification error rate is maximized. The error rate is composed of the false alarm probability and the missed alarm probability.

[0064] The final formula is as follows:

[0065]

[0066] Where N represents the number of channels; n represents the channel number; Δω kw P represents the detection error of UAV k against interceptor w; kk′ λ represents the power allocated for communication between the k-th UAV and the k′-th UAV. n This represents the wavelength corresponding to the assigned channel n; g kn Indicates whether the nth channel is assigned to the kth UAV for radar detection; β represents the channel gain coefficient; u k This represents the location information of drone k; ω w This indicates the location information of the interceptor ω; This represents the maximum resource limit under the condition of maximum interception error rate; σ 2 This represents the noise variance.

[0067] The performance indicators, mainly composed of the combined performance of UAV swarm communication and target detection, are as follows:

[0068]

[0069] (Δω kw ) 2 ≤δ

[0070]

[0071] Where K is the number of drones in the swarm; k is the individual drone number in the swarm; k′ is the communication target of drone k; N represents the number of channels; n represents the channel number; P kk′ σ represents the power allocated for communication between the k-th drone and the k′-th drone; 2 h represents the noise variance. kk′n This represents the channel gain allocated to the communication between the k-th UAV and the k′-th UAV using channel n. This represents the probability that the most recently intercepted drone is true when drone k is communicating. This represents the probability that the most recently intercepted device was detected as a fake. kk′n This indicates whether the nth channel is allocated for communication between the kth and k′th drones; 0 indicates no allocation, and 1 indicates allocation. kn Indicates whether the nth channel is allocated to the kth UAV for interception detection; δ represents the radar detection threshold; Δω kw λ represents the detection error of UAV k against interceptor w; n P represents the wavelength corresponding to the assigned channel n; kk′r This represents the received power when the k-th drone communicates with the k′-th drone; L represents the square of the distance from which drone k detected the nearest interceptor; kk′G represents the system loss coefficient for communication between the k-th UAV and the k′-th UAV; kk′t and G kk′r These represent the transmit gain and receive gain, respectively. β represents the channel gain coefficient; u k This represents the location information of drone k; ω w This indicates the location information of the interceptor ω; This represents the maximum resource limit under the condition of the highest interception error rate;

[0072] 3) Based on the system, a resource scheduling method is adopted. Resource scheduling refers to the rational allocation of system resources for the two main tasks of inter-UAV swarm communication and target detection. Resource scheduling mainly involves three aspects: system resources, system resource allocation methods, and resource constraints.

[0073] System resources:

[0074] like Figure 2 The diagram shows the system resource map. System resources mainly include the operating center frequency, bandwidth, and power. Bandwidth refers to the bandwidth at the corresponding operating center frequency. Furthermore, each bandwidth is divided into multiple channels, and each channel cannot be repeatedly allocated. Additionally, power resources limit the energy allocated to each UAV for radar detection and wireless communication. Finally, the free combination of these three resources constitutes a resource block.

[0075] Allocation method:

[0076] like Figure 3 The diagram shows the system resource allocation method. The system resource allocation method means that each communication segment and target detection segment within the UAV swarm will be allocated a resource block.

[0077] Resource limitations:

[0078] Based on the most basic communication requirements, covert communication has a minimum communication rate limit;

[0079]

[0080] Where N represents the number of channels; n represents the channel number; P kk′ h represents the power allocated for communication between the k-th drone and the k′-th drone. kk′n The channel gain σ represents whether the nth channel is allocated to communication between the kth and k′ drones; 2 Indicates the noise figure; This indicates the threshold for the most basic communication requirements of a drone.

[0081] Based on the most basic detection requirements, radar detection has minimal detection resource constraints.

[0082]

[0083] Where N represents the number of channels; n represents the channel number; T p Indicates pulse width; υ represents the radar cross-section (RCS); P k h represents the power allocated to the k-th UAV during its reconnaissance mission. kn σ represents the channel gain when the nth channel is allocated to the kth UAV for detection; 2 Indicates the noise figure; This indicates the threshold for the most basic radar detection requirements of drones.

[0084] 4) To facilitate resource allocation in terms of operating frequency and bandwidth, a reinforcement learning-based resource allocation model is constructed, and resource optimization is achieved by optimizing the reward function. The reinforcement learning-based resource allocation model includes environmental state, agent behavior, and reward.

[0085] Environmental state: ε [t] It indicates the current environmental state and is composed of the communication links of the UAV swarm and the channels allocated for target detection.

[0086]

[0087] in, Represents a real dataset; K represents the number of drone swarms; M represents... [t] ={m kk′ ,k=1,2,…,K} represents the channel allocated for the communication link between drone swarms. m kk′ E represents the channel allocated for communication between drone k and drone k′. [t] ={e k ,k=1,2,…,K} represents the channel assigned to each UAV when performing target detection. k This represents the channel assigned to the UAV k when it performs target detection.

[0088] Agent behavior: in state ε [t] Let A be all possible behaviors. [t] .

[0089]

[0090] The number of actions is F. [t] The f-th action is defined as This represents the o-th channel at the j-th operating frequency. [t] and O [t] Let A and B represent the allocatable working center frequency and channel in the current state, respectively. The agent selects an action from the action set based on the current state and reward. Note that, considering channel interference, action set A... [t]Actions that use the same channel as the previous actions are not included.

[0091] Reward: Based on the current resource allocation strategy ε [t] The reward is defined as a performance indicator of the covert communication of the drone swarm system, and our goal is to maximize the reward.

[0092] 5) To facilitate power allocation, a resource allocation process based on the water injection method is constructed.

[0093] First, assuming the target detection power is a fixed value, calculate the communication power between the drone swarm:

[0094]

[0095] Among them, P kk′ f represents the power allocated for communication between drone k and drone k′. k This represents the probability of identifying a drone as real, assuming the nearest interceptor to drone k is real. kk′S This represents the number of channels allocated for communication between UAV k and UAV k′ when the operating center frequency is S. L1 represents the Lagrange multiplier corresponding to the communication power. r This indicates that the target detection power is a fixed value at this time. This represents the total power limit of drone k.

[0096] Then, based on the obtained communication power, calculate the target detection power:

[0097]

[0098] Among them, P k f represents the power allocated to UAV k for target detection. kk′ This represents the communication rate between drone k and drone k′. kS This represents the number of channels allocated to UAV k for target detection when the operating center frequency is S. L2 represents the Lagrange multiplier corresponding to the target detection power. kk′ This represents the power allocated for communication between drone k and drone k′. This represents the total power limit of drone k.

[0099] Specific parameters are shown in Tables 1 and 2. The number of UAV clusters, K1, is 3; the number of interceptors is 3; the total number of channels is 21; the wavelength is 1; and the algorithm used is the DQN algorithm.

[0100] Table 1 Simulation Parameters

[0101] Number of drone swarms 3 Number of intercepted aircraft 3 Number of channels 21 Communication transmission power (used for reinforcement learning) 10000W Radar transmit power (used for reinforcement learning) 100KW System noise temperature 290K Boltzmann constant <![CDATA[1.38×10 -23 ]]> Number of wavelengths 1 Power range limitation 200000000W

[0102] Table 2 Parameters of Spectrum Resource Allocation Algorithm

[0103]

[0104] Specific results are as follows Figure 4 As shown, from Figure 4 The graph clearly shows that the DQN algorithm is used for reasonable wavelength and channel resource allocation. The x-axis represents the number of rounds, or iterations; the y-axis represents the reward. The graph shows that the total reward value is not high in the early stages of learning. As the number of iterations increases, the total reward value tends to reach its maximum and then stabilizes. Based on the DQN algorithm, a water-filling method is used to reasonably allocate power.

[0105] like Figure 5 The figure shows the test results of this invention under varying channel numbers. The X-axis represents the number of channels, and the y-axis represents the total reward. It is clear from the figure that this invention is significantly superior to other algorithms.

[0106] like Figure 6 The figure shows the test results of this invention under varying numbers of drones. The X-axis represents the number of users, and the y-axis represents the total reward. It is clear from the figure that this invention is significantly superior to other algorithms.

[0107] The present invention has been described above with reference to embodiments. It is understood that not all implementations can be exhaustively described in practice. Those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are protected by the present invention.

Claims

1. A method for integrating radar and communication in a UAV swarm based on resource scheduling, characterized in that, Includes the following steps: Step 1: Based on the performance requirements of covert communication and detection of UAV swarms, construct an integrated radar and communication system for UAV swarms; Step 2: Process the system using resource scheduling methods; Step 3: In order to effectively allocate system resources, construct an optimized resource allocation method based on a combination of reinforcement learning and water injection method; The radar-communication integrated system of the UAV swarm described in step 1 is as follows: it consists of several UAVs equipped with radar-communication integrated equipment; under the condition that the enemy interceptor cannot effectively intercept them, the several UAVs conduct covert communication with each other based on the detection, identification and location of the interceptor; the several UAVs rationally allocate resources for communication and target detection through the radar-communication integrated equipment. The conditions under which the enemy interceptor aircraft cannot effectively intercept are as follows: By optimizing the design of radar and communication resources, the enemy interceptor aircraft is unable to effectively intercept the wireless communication between the drone swarm. The enemy interceptor aircraft uses hypothesis testing to identify the intercepted signals. By optimizing resource allocation, the identification error rate is maximized. The error rate is composed of the false alarm probability and the missed alarm probability. The final error rate formula is as follows: Where N represents the number of channels; n represents the channel number; Δω kw P represents the detection error of UAV k against interceptor w; kk' λ represents the power allocated for communication between the k-th drone and the k'-th drone; n This represents the wavelength corresponding to the assigned channel n; g kn Indicates whether the nth channel is assigned to the kth UAV for radar detection; β represents the channel gain coefficient; u k This represents the location information of drone k; ω w This indicates the location information of the interceptor ω; This represents the maximum resource limit under the condition of maximum interception error rate; σ 2 This represents the noise variance.

2. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 1, characterized in that, The covert communication performance indicators of the UAV swarm system described in step 1 are as follows: (See kw ) 2 ≤δ Where K is the number of drones in the swarm; k is the individual drone number in the swarm; k' is the communication target of drone k; N represents the number of channels; n represents the channel number; P kk' σ represents the power allocated for communication between the k-th drone and the k'-th drone; 2 h represents the noise variance. kk'n This represents the channel gain allocated to the communication between the k-th UAV and the k'-th UAV using channel n. This represents the probability that the most recently intercepted drone is true when drone k is communicating. Indicates the probability that the most recently intercepted device is a fake; g kk'n This indicates whether the nth channel is allocated for communication between the kth and k'th drones; 0 indicates no allocation, and 1 indicates allocation. kn Indicates whether the nth channel is allocated to the kth UAV for interception detection; δ represents the radar detection threshold; Δω kw λ represents the detection error of UAV k against interceptor w; n P represents the wavelength corresponding to the assigned channel n; kk'r This represents the received power when the k-th drone communicates with the k'-th drone; L represents the square of the distance from which drone k detected the nearest interceptor; kk' G represents the system loss coefficient for communication between the k-th drone and the k'-th drone; kk't and G kk'r These represent the transmit gain and receive gain, respectively; β represents the channel gain coefficient; u k This represents the location information of drone k; ω w This indicates the location information of the interceptor ω; This represents the maximum resource limit under the condition of the highest interception error rate.

3. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 1, characterized in that, The system resources and their allocation methods corresponding to the resource scheduling method in step 2 are as follows: Resources include operating center frequency, bandwidth, and power. Bandwidth refers to the bandwidth at the corresponding operating center frequency. Secondly, each bandwidth is divided into multiple channels, and each channel cannot be repeatedly allocated. In addition, power resources limit the energy of each UAV used for radar detection and wireless communication. Finally, the free combination of the three resources constitutes a resource block. The system resource allocation method is as follows: communication between UAVs and target detection are each allocated a resource block.

4. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 1, characterized in that, The limitations of the resource scheduling method described in step 2 are as follows: Based on the most basic communication requirements, covert communication has a minimum communication rate limit: Where N represents the number of channels; n represents the channel number; P kk' h represents the power allocated for communication between the k-th drone and the k'-th drone. kk'n The channel gain σ represents whether the nth channel is allocated to communication between the kth and k'th drones; 2 Indicates the noise figure; This indicates the threshold for the most basic communication requirements of a drone; Based on the most basic detection requirements, radar detection has minimal detection resource constraints; Where N represents the number of channels; n represents the channel number; T p Indicates pulse width; υ represents the radar cross-section (RCS); P k h represents the power allocated to the k-th UAV during its reconnaissance mission. kn σ represents the channel gain when the nth channel is allocated to the kth UAV for detection; 2 Indicates the noise figure; This indicates the threshold for the most basic radar detection requirements of drones.

5. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 1, characterized in that, Step 3 allocates the operating center frequency and corresponding bandwidth using reinforcement learning, and allocates power using a water-filling method. Specifically, it includes the following steps: Step 3.1: Construct an environment model for reinforcement learning, including environment state, agent behavior, and reward; Step 3.2: Construct a model for the water injection method; Step 3.3: Based on the obtained communication power, calculate the target detection power.

6. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 5, characterized in that, Step 3.1, which involves constructing the environment model for reinforcement learning, specifically includes: Environmental state: ε [t] It indicates the current environmental state and is composed of the communication links of the UAV swarm and the channels allocated for target detection. in, Represents a real dataset; K represents the number of drone swarms; M represents... [t] ={m kk' {k = 1, 2, ..., K} represents the channels allocated for communication links between UAV swarms; m kk' E represents the channel allocated for communication between drone k and drone k'. [t] ={e k {k = 1, 2, ..., K} represents the channel allocated to each UAV for target detection; e k This represents the channel assigned to UAV k when it performs target detection; Agent behavior: in state ε [t] Let A be all possible behaviors. [t] ; The number of actions is F. [t] The f-th action is defined as J represents the o-th channel at the j-th operating frequency; [t] and O [t] These represent the allocatable working center frequency and channel in the current state; the agent selects an action from the action set based on the current state and reward; Action set A [t] This does not include actions that use the same channel as the previous actions; Reward: Based on the current resource allocation strategy ε [t] The reward is defined as a performance indicator of the covert communication of the drone swarm system, and the goal is to maximize the reward.

7. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 5, characterized in that, Step 3.2 specifically includes: First, assuming the target detection power is a fixed value, calculate the communication power between the drone swarm: Among them, P kk' f represents the power allocated for communication between drone k and drone k'. k This represents the probability of identifying a drone as real, assuming the nearest interceptor to drone k is real; n kk'S L1 represents the number of channels allocated for communication between UAV k and UAV k' when the working center frequency is S; L1 represents the Lagrange coefficients corresponding to the communication power; P r This indicates that the target detection power is a fixed value at this time; This represents the total power limit of drone k.

8. The method for integrating radar and communication in UAV swarms based on resource scheduling according to claim 5, characterized in that, In step 3.3, based on the obtained communication power, the target detection power is calculated: Among them, P k f represents the power allocated to UAV k for target detection. kk' This represents the communication rate between drone k and drone k'; n kS L2 represents the number of channels allocated to UAV k for target detection when the working center frequency is S; L2 represents the Lagrange multiplier corresponding to the target detection power; P kk' This represents the power allocated for communication between drone k and drone k'. This represents the total power limit of drone k.

Images