System cost function optimization-based flight path optimization and resource allocation method for unmanned aerial vehicle communication and induction integrated system
By modeling the system cost function in the integrated drone synesthesia system, optimizing the drone communication perception precoding and scheduling strategy, the resource competition problem is solved, the system performance is improved, and communication perception in complex scenarios is met need.
Patent Information
- Application Number
- CN202410910920.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-09
- Publication Date
- 2024-11-01
AI Technical Summary
The multi-dimensional resource competition problems such as spectrum, time slot and power of communication and perception functions in the integrated communication and perception system of drone synesthesia system are difficult to achieve performance optimization of communication and perception integration system in complex scenarios.
For system scenarios of single multi-antenna drones, multiple users and perceived targets, the system cost function is modeled to realize the joint optimization of the drone communication perception precoding, communication perception scheduling strategy and drone trajectory, using dual Q network and greedy algorithm optimize the design of flight trajectory and resource allocation strategies.
The combined optimization of drone communication perception precoding, communication perception scheduling strategy and drone trajectory has been realized, which improves system performance and meets the performance requirements of user communication and target perception.
Smart Images

Figure CN118890116A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of wireless communications, and relates to a flight trajectory optimization and resource allocation method for an unmanned aerial vehicle synaesthesia integrated system based on system cost function optimization. Background Art
[0002] Drones have been widely used in wireless communication and target perception systems due to their high flexibility, low cost and easy deployment. By deploying communication perception modules on drones and combining drones with communication perception integration technology, drone-assisted user communication and target perception can be achieved simultaneously, improving information transmission and target perception performance. However, the drone interawareness integration system has the problem of competition for multi-dimensional resources such as spectrum, time slots and power in communication and perception functions. How to design the drone flight trajectory and resource allocation method to achieve performance optimization of the communication perception system is an urgent problem to be solved.
[0003] At present, there are existing literatures studying the problem of integrated communication and perception of UAVs. For example, based on the problem of tight spectrum resources of communication and perception systems, some literatures design beam sharing schemes to improve the collaborative perception performance of UAVs; some literatures propose a UAV-assisted communication perception system that integrates periodic perception and communication mechanisms, and maximizes the user's achievable rate by jointly optimizing the UAV trajectory, transmission precoding and periodic sensing perception time slots. However, existing research rarely considers the joint optimization of communication and perception system performance, as well as the UAV flight trajectory, communication perception precoding and communication perception scheduling design issues, making it difficult to achieve performance optimization of the communication perception integrated system in complex scenarios. Summary of the invention
[0004] In view of this, the purpose of the present invention is to provide a flight trajectory optimization and resource allocation method for a UAV synaesthesia integrated system based on system cost function optimization. For a system scenario including a single multi-antenna UAV, J users and K sensing targets, the system cost function is modeled as the optimization target to achieve the joint optimization of UAV communication perception precoding, communication perception scheduling strategy and UAV trajectory.
[0005] In order to achieve the above object, the present invention provides the following technical solutions:
[0006] A method for optimizing the flight trajectory and allocating resources of a UAV synaesthesia integrated system based on system cost function optimization is proposed for a system scenario with a single multi-antenna UAV, J users and K sensing targets. The method specifically includes the following steps:
[0007] S1: Modeling UAV synaesthesia integrated system;
[0008] S2: Modeling UAV communication channel model;
[0009] S3: Modeling UAV perception channel model;
[0010] S4: Modeling UAV communication perception scheduling variables;
[0011] S5: Modeling user queue length;
[0012] S6: Modeling UAV target estimation Cramer-Rao bound;
[0013] S7: Modeling UAV energy consumption;
[0014] S8: Modeling system cost function;
[0015] S9: Modeling communication perception constraints and UAV trajectory constraints;
[0016] S10: Determine the UAV flight trajectory and resource allocation strategy based on system cost function optimization.
[0017] Further, in step S1, a UAV interawareness integrated system model is built, which specifically includes: the system includes a single multi-antenna UAV, J users and K sensing targets, and the number of UAV antennas is M; the UAV has a built-in communication sensing module, which can communicate with the user and send data packets, and can also send radar signals to estimate the target state; the system time is divided into T equal-length time slots, and the length of each time slot is τ; the UAV uses frequency division multiple access technology for communication and target estimation, let B, B s They represent the bandwidth of the communication channel and the perception channel respectively; assuming that the UAV's flight altitude is fixed at H and its maximum flight speed is v max , let q(t)=[x u (t),y u (t)] T represents the position of the drone in time slot t, 1≤t≤T, let ω j =[x j ,y j ] T represents the position of the jth user, 1≤j≤J, let represents the position of the kth target.
[0018] Further, in step S2, a UAV communication channel model is built, specifically including: let h j (t)∈C M×1 represents the communication channel gain vector between the UAV and the jth user in time slot t, which is modeled as:
[0019]
[0020] Where κ represents the attenuation factor of the non-line-of-sight NLoS channel, β0 represents the channel power gain at a reference distance of 1m, and d j(t) represents the distance between the UAV and the jth user, which can be modeled as They represent the line-of-sight LoS link probability and NLoS link probability between the UAV and the jth user in time slot t, that is, represents the angle between the drone and the jth user, represents the array steering vector between the UAV and the jth user in time slot t, which is modeled as The array steering vector representing the NLoS is modeled as a complex Gaussian random vector with zero mean and unit covariance matrix.
[0021] Further, in step S3, a drone perception channel model is modeled, specifically including: It represents the perception channel gain matrix between the UAV and the kth target in time slot t, which can be modeled as:
[0022]
[0023] Among them, μ k represents the complex reflection coefficient, and They represent the LoS link and NLoS link probabilities between the UAV and the kth target, namely, represents the angle between the UAV and the kth target, represents the array steering matrix between the UAV and the kth user, that is, where g k (t)∈C M×1 represents the array steering vector between the UAV and the kth user, which can be modeled as Represents the array steering matrix of NLoS, modeling a complex Gaussian random variable with zero mean and unit variance.
[0024] Further, in step S4, the UAV communication perception scheduling variables are modeled, specifically including: let α j (t) represents the scheduling variable between the UAV and the jth user in time slot t, α j (t) = 1 means that the UAV communicates with the jth user in the t-th time slot, otherwise, α j (t) = 0; let represents the perception scheduling variable between the UAV and the kth target in time slot t, Indicates that the UAV perceives the kth target in time slot t, and vice versa.
[0025] Further, in step S5, modeling the user queue length specifically includes: setting Q j (t) represents the queue length of the jth user in time slot t, which can be modeled as:
[0026]
[0027] Among them, R j (t) represents the communication rate of the jth user in time slot t, which is modeled as:
[0028]
[0029] H c (t)∈C M×J represents the channel gain matrix between the user and the UAV in time slot t, W c (t)∈C M×J represents the communication precoding matrix of time slot t, σ 2 represents the noise power, λ j (t) represents the amount of data randomly arriving at the jth user in time slot t, which follows the Poisson distribution. Indicates the maximum queue length of the j-th user.
[0030] Further, in step S6, the UAV target is modeled to estimate the Cramer-Rao bound, specifically including: after the UAV receives the echo signal from the kth target, it estimates the target angle make represents the t time slot estimate The Cramer-Rao bound of can be expressed as:
[0031]
[0032] in, Represents G k (t) Yes The partial derivative of
[0033] Further, in step S7, the energy consumption of the drone is modeled, specifically including: let E f (t) represents the flight energy consumption of the UAV in time slot t, which can be modeled as: Among them, δ f (t)∈{0,1} represents the flight variable of the UAV at time slot t. If δ f (t) = 1, the UAV is in flight at time slot t, otherwise, δ f (t) = 0, P0, P0′ represent the constants of blade profile power and induced power of the UAV in hovering state, U tip is the tip speed of the rotor blade, v0 is the average rotor induced speed in the hovering state, and v t is the flight speed of the drone in time slot t, ξ d and r are the fuselage drag ratio and rotor reliability, ρ and S are the air density and rotor disc area, respectively; let E h(t) represents the hovering energy consumption of the UAV in time slot t, which is modeled as: E h (t) = δ h (t)(P0+P0′)τ, where δ h (t)∈{0,1} represents the hovering variable of the drone in time slot t. If δ h (t) = 1, the UAV is in hovering state at time slot t, otherwise, δ h (t) = 0; let E c (t), E s (t) represent the communication energy consumption and perception energy consumption of the UAV, which can be modeled as E c (t) = τTr(W c (t)(W c (t)) H ),
[0034] Further, in step S8, the system cost function is modeled, specifically including: taking into account the user queue length and the drone perception performance, the system cost function is modeled as a weighted function of the longest user queue length and the maximum target angle estimated Cramer-Rao bound. Let η(t) represent the system cost function of time slot t, which can be modeled as Among them, β1 and β2 represent weight factors respectively.
[0035] Furthermore, in step S9, the communication perception constraint and the drone trajectory constraint are modeled, specifically including: any user must communicate at least once within the system time. This constraint can be expressed as: In any time slot, the drone can communicate with at most N users simultaneously. This constraint can be expressed as: Any target is sensed at least once within the system time. This constraint can be expressed as: In any time slot, the drone can sense at most one target. This constraint can be expressed as: The user transmission rate should meet the minimum communication rate threshold. This constraint can be expressed as: in represents the minimum communication rate threshold of the jth user; the Cramer-Rao bound of the drone's estimated target should be less than a threshold to meet the perception performance. This constraint can be expressed as: where γ th (t) represents the Cramer-Rao bound threshold of the drone’s estimated target; the total power perceived by the drone’s communication does not exceed its maximum transmit power. This constraint can be expressed as: Where P max represents the maximum transmission power of the UAV; the total energy consumption of the UAV should not exceed the total energy. This constraint can be expressed as: Where E maxrepresents the total energy of the drone; the flight distance of the drone between two consecutive time slots should satisfy the distance constraint, which can be expressed as:
[0036] Further, in step S10, the flight trajectory of the UAV and the resource allocation strategy are determined based on the system cost function optimization, specifically including: the modeling system performance metric is: The UAV trajectory and resource allocation problem is modeled as a Markov decision process MDP problem, modeling the system state, action and benefit function; the system state in time slot t is in, represents the length of the user queue in time slot t, E r (t) represents the remaining energy of the UAV in time slot t, which can be expressed as: Let a(t) represent the action space of the drone in time slot t, which can be expressed as: Where z represents the flight distance of the drone in a time slot, z = v max τ; let r(t) represent the reward function of time slot t, and model it as follows: r(t) = -η(t)-h1+h2+h3, where h1 represents the penalty factor for the exhaustion of the drone battery energy, h2 represents the reward factor for all users to communicate at least once within the system time, and h3 represents the reward factor for all targets to be perceived at least once within the system time. The flight trajectory and resource allocation strategy of the drone are optimized based on the dual Q network DDQN framework and the greedy algorithm. Specifically, given the state and action of the MDP, the communication perception scheduling strategy of the drone is determined based on the greedy algorithm. Determine the UAV communication and perception precoding scheme W based on convex optimization method c (t), W s (t); UAV communication perception scheduling strategy based on given time slot And the precoding scheme W c (t), W s (t), Determine the flight trajectory of the drone based on long-term reward optimization
[0037] The beneficial effects of the present invention are as follows: the present invention targets a system scenario including a single multi-antenna UAV, multiple users and multiple perception targets, models the system cost function as an optimization target, and realizes the joint optimization of UAV communication perception precoding, communication perception scheduling strategy and UAV trajectory.
[0038] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:
[0040] Figure 1 This is a schematic diagram of the scene of the UAV synaesthesia integration system of the present invention;
[0041] Figure 2 The present invention is a flow chart of the flight trajectory optimization and resource allocation method of the unmanned aerial vehicle synaesthesia integrated system based on system cost function optimization. DETAILED DESCRIPTION
[0042] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.
[0043] See also Figure 1-2 , which is a flight trajectory optimization and resource allocation method for a UAV synaesthesia integrated system based on system cost function optimization provided by the present invention. For a system scenario including a single multi-antenna UAV, multiple users and multiple perception targets, the modeling system cost function is used as the optimization target to achieve the joint optimization of UAV communication perception precoding, communication perception scheduling strategy and UAV trajectory.
[0044] Figure 1 Schematic diagram of the UAV synaesthesia integrated system scenario of the present invention. Figure 1 As shown in the figure, there is a single multi-antenna UAV, J users and K sensing targets in the scene, and the number of transmitting antennas of the UAV is M. The UAV is equipped with airborne communication and radar equipment, and can send data to communication users based on the downlink communication link. At the same time, the UAV can also send radar signals through the airborne radar and receive target echo signals to perceive target information.
[0045] Figure 2 FIG. 1 is a flow chart of the method for optimizing the flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system based on the optimization of the system cost function of the present invention. Figure 2 As shown, the method specifically comprises the following steps:
[0046] 1) Modeling the UAV synaesthesia integrated system model;
[0047] The UAV interawareness integrated system model includes: the system includes a single multi-antenna UAV, J users and K sensing targets, and the number of UAV antennas is M; the UAV has a built-in communication and perception module, which can communicate with users and send data packets, and can also send radar signals to estimate the target state; the system time is divided into T equal-length time slots, and the length of each time slot is τ; the UAV uses frequency division multiple access technology for communication and target estimation, let B, B s They represent the bandwidth of the communication channel and the perception channel respectively; assuming that the UAV's flight altitude is fixed at H and its maximum flight speed is v max , let q(t)=[x u (t),y u (t)] T represents the position of the drone in time slot t, 1≤t≤T, let ω j =[x j ,y j ] T represents the position of the jth user, 1≤j≤J, let represents the position of the kth target.
[0048] 2) Modeling the UAV communication channel model;
[0049] Modeling the UAV communication channel model includes: let h j (t)∈C M×1 represents the communication channel gain vector between the UAV and the jth user in time slot t, which is modeled as:
[0050]
[0051] Where κ represents the attenuation factor of the non-line-of-sight NLoS channel, β0 represents the channel power gain at a reference distance of 1m, and d j (t) represents the distance between the UAV and the jth user, which can be modeled as They represent the line-of-sight LoS link probability and NLoS link probability between the UAV and the jth user in time slot t, that is, represents the angle between the drone and the jth user, represents the array steering vector between the UAV and the jth user in time slot t, which is modeled as The array steering vector representing the NLoS is modeled as a complex Gaussian random vector with zero mean and unit covariance matrix.
[0052] 3) Modeling the UAV perception channel model;
[0053] Modeling the UAV perception channel model includes: It represents the perception channel gain matrix between the UAV and the kth target in time slot t, which can be modeled as:
[0054]
[0055] Among them, μ k represents the complex reflection coefficient, and They represent the LoS link and NLoS link probabilities between the UAV and the kth target, namely, represents the angle between the UAV and the kth target, represents the array steering matrix between the UAV and the kth user, that is, where g k (t)∈C M×1 represents the array steering vector between the UAV and the kth user, which can be modeled as Represents the array steering matrix of NLoS, modeling a complex Gaussian random variable with zero mean and unit variance.
[0056] 4) Modeling UAV communication perception scheduling variables;
[0057] Modeling UAV communication perception scheduling variables includes: Let α j (t) represents the scheduling variable between the UAV and the jth user in time slot t, α j (t) = 1 means that the UAV communicates with the jth user in the t-th time slot, otherwise, α j (t) = 0; let represents the perception scheduling variable between the UAV and the kth target in time slot t, Indicates that the UAV perceives the kth target in time slot t, and vice versa.
[0058] 5) Modeling user queue length;
[0059] Modeling the length of user queues includes: Let Q j (t) represents the queue length of the jth user in time slot t, which can be modeled as:
[0060]
[0061] Among them, R j (t) represents the communication rate of the jth user in time slot t, which is modeled as:
[0062]
[0063] H c (t)∈C M×J represents the channel gain matrix between the user and the UAV in time slot t, W c (t)∈C M×J represents the communication precoding matrix of time slot t, σ 2 represents the noise power, λ j(t) represents the amount of data randomly arriving at the jth user in time slot t, which follows the Poisson distribution. Indicates the maximum queue length of the j-th user.
[0064] 6) Modeling UAV target estimation Cramer-Rao bound;
[0065] Modeling UAV target estimation Cramer-Rao bound includes: After the UAV receives the echo signal from the kth target, it estimates the target angle make represents the t time slot estimate The Cramer-Rao bound of can be expressed as:
[0066]
[0067] in, Represents G k (t) Yes The partial derivative of
[0068] 7) Modeling drone energy consumption;
[0069] Modeling the energy consumption of the drone includes: f (t) represents the flight energy consumption of the UAV in time slot t, which can be modeled as: Among them, δ f (t)∈{0,1} represents the flight variable of the UAV at time slot t. If δ f (t) = 1, the UAV is in flight at time slot t, otherwise, δ f (t) = 0, P0, P0′ represent the constants of blade profile power and induced power of the UAV in hovering state, U tip is the tip speed of the rotor blade, v0 is the average rotor induced speed in the hovering state, and v t is the flight speed of the drone in time slot t, ξ d and r are the fuselage drag ratio and rotor reliability, ρ and S are the air density and rotor disc area, respectively; let E h (t) represents the hovering energy consumption of the UAV in time slot t, which is modeled as: E h (t) = δ h (t)(P0+P0′)τ, where δ h (t)∈{0,1} represents the hovering variable of the drone in time slot t. If δ h (t) = 1, the UAV is in hovering state at time slot t, otherwise, δ h (t) = 0; let E c (t), E s (t) represent the communication energy consumption and perception energy consumption of the UAV, which can be modeled as Ec (t) = τTr(W c (t)(W c (t)) H ),
[0070] 8) Modeling system cost function;
[0071] The modeling system cost function includes: taking into account the user queue length and the drone perception performance, the modeling system cost function is a weighted function of the longest user queue length and the maximum target angle estimated Cramer-Rao bound. Let η(t) represent the system cost function of time slot t, which can be modeled as Among them, β1 and β2 represent weight factors respectively.
[0072] 9) Modeling communication perception constraints and drone trajectory constraints;
[0073] Modeling communication perception constraints and drone trajectory constraints include: any user communicates at least once within the system time. This constraint can be expressed as: In any time slot, the drone can communicate with at most N users simultaneously. This constraint can be expressed as: Any target is sensed at least once within the system time. This constraint can be expressed as: In any time slot, the drone can sense at most one target. This constraint can be expressed as: The user transmission rate should meet the minimum communication rate threshold. This constraint can be expressed as: in represents the minimum communication rate threshold of the jth user; the Cramer-Rao bound of the drone's estimated target should be less than a threshold to meet the perception performance. This constraint can be expressed as: where γ th (t) represents the Cramer-Rao bound threshold of the drone’s estimated target; the total power perceived by the drone’s communication does not exceed its maximum transmit power. This constraint can be expressed as: Where P max represents the maximum transmission power of the UAV; the total energy consumption of the UAV should not exceed the total energy. This constraint can be expressed as: Where E max represents the total energy of the drone; the flight distance of the drone between two consecutive time slots should satisfy the distance constraint, which can be expressed as:
[0074] 10) Determine the UAV flight trajectory and resource allocation strategy based on system cost function optimization;
[0075] Determining the flight trajectory of the UAV and the resource allocation strategy based on the optimization of the system cost function includes: Modeling system performance metrics are: The UAV trajectory and resource allocation problem is modeled as a Markov decision process MDP problem, modeling the system state, action and benefit function; the system state in time slot t is in, represents the length of the user queue in time slot t, E r (t) represents the remaining energy of the UAV in time slot t, which can be expressed as: Let a(t) represent the action space of the drone in time slot t, which can be expressed as: Where z represents the flight distance of the drone in a time slot, z = v max τ; let r(t) represent the reward function of time slot t, and model it as follows: r(t) = -η(t)-h1+h2+h3, where h1 represents the penalty factor for the exhaustion of the drone battery energy, h2 represents the reward factor for all users to communicate at least once within the system time, and h3 represents the reward factor for all targets to be perceived at least once within the system time. The flight trajectory and resource allocation strategy of the drone are optimized based on the dual Q network DDQN framework and the greedy algorithm. Specifically, given the state and action of the MDP, the communication perception scheduling strategy of the drone is determined based on the greedy algorithm. Determine the UAV communication and perception precoding scheme W based on convex optimization method c (t), W s (t); UAV communication perception scheduling strategy based on given time slot And the precoding scheme W c (t), W s (t), Determine the flight trajectory of the drone based on long-term reward optimization
[0076]
[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.
Claims
1. A method for optimizing flight trajectory and allocating resources for an unmanned aerial vehicle synaesthesia integrated system based on system cost function optimization, characterized in that: include: S1: Modeling UAV synaesthesia integrated system; S2: Modeling UAV communication channel model; S3: Modeling UAV perception channel model; S4: Modeling UAV communication perception scheduling variables; S5: Modeling user queue length; S6: Modeling UAV target estimation Cramer-Rao bound; S7: Modeling UAV energy consumption; S8: Modeling system cost function; S9: Modeling communication perception constraints and UAV trajectory constraints; S10: Determine the UAV flight trajectory and resource allocation strategy based on system cost function optimization.
2. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 1 is characterized in that: In step S1, a UAV interawareness integrated system model is built, which specifically includes: the system includes a single multi-antenna UAV, J users and K sensing targets, and the number of UAV antennas is M; the UAV has a built-in communication sensing module, which can communicate with the user and send data packets, and can also send radar signals to estimate the target state; the system time is divided into T equal-length time slots, and the length of each time slot is τ; the UAV uses frequency division multiple access technology for communication and target estimation, let B, B s They represent the bandwidth of the communication channel and the perception channel respectively; assuming that the UAV's flight altitude is fixed at H and its maximum flight speed is v max , let q(t)=[x u (t),y u (t)] T represents the position of the drone in time slot t, 1≤t≤T, let ω j =[x j ,y j ] T represents the position of the jth user, 1≤j≤J, let represents the position of the kth target.
3. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 2 is characterized in that: In step S2, a UAV communication channel model is built, which specifically includes: let h j (t)∈C M×1 represents the communication channel gain vector between the UAV and the jth user in time slot t, which is modeled as: Where κ represents the attenuation factor of the non-line-of-sight NLoS channel, β0 represents the channel power gain at a reference distance of 1m, and d j (t) represents the distance between the UAV and the jth user, which can be modeled as They represent the line-of-sight LoS link probability and NLoS link probability between the UAV and the jth user in time slot t, that is, represents the angle between the drone and the jth user, represents the array steering vector between the UAV and the jth user in time slot t, which is modeled as The array steering vector representing the NLoS is modeled as a complex Gaussian random vector with zero mean and unit covariance matrix.
4. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 3 is characterized in that: In step S3, the UAV perception channel model is built, specifically including: It represents the perception channel gain matrix between the UAV and the kth target in time slot t, which can be modeled as: Among them, μ k represents the complex reflection coefficient, and They represent the LoS link and NLoS link probabilities between the UAV and the kth target, namely, represents the angle between the UAV and the kth target, represents the array steering matrix between the UAV and the kth user, that is, where g k (t)∈C M×1 represents the array steering vector between the UAV and the kth user, which can be modeled as Represents the array steering matrix of NLoS, modeling a complex Gaussian random variable with zero mean and unit variance.
5. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 4 is characterized in that: In step S4, the UAV communication perception scheduling variables are modeled, specifically including: let α j (t) represents the scheduling variable between the UAV and the jth user in time slot t, α j (t) = 1 means that the UAV communicates with the jth user in the t-th time slot, otherwise, α j (t) = 0; let represents the perception scheduling variable between the UAV and the kth target in time slot t, Indicates that the UAV perceives the kth target in time slot t, and vice versa.
6. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 5 is characterized in that: In step S5, modeling the user queue length includes: setting Q j (t) represents the queue length of the jth user in time slot t, which can be modeled as: Among them, R j (t) represents the communication rate of the jth user in time slot t, which is modeled as: H c (t)∈C M×J represents the channel gain matrix between the user and the UAV in time slot t, W c (t)∈C M×J represents the communication precoding matrix of time slot t, σ 2 represents the noise power, λ j (t) represents the amount of data randomly arriving at the jth user in time slot t, which follows the Poisson distribution. Indicates the maximum queue length of the j-th user.
7. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 6 is characterized in that: In step S6, the UAV target is modeled to estimate the Cramer-Rao bound, which specifically includes: after the UAV receives the echo signal from the kth target, it estimates the target angle make represents the t time slot estimate The Cramer-Rao bound of can be expressed as: in, Represents G k (t) Yes The partial derivative of 8. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 7, characterized in that: In step S7, the energy consumption of the drone is modeled, specifically including: let E f (t) represents the flight energy consumption of the UAV in time slot t, which can be modeled as: Among them, δ f (t)∈{0,1} represents the flight variable of the UAV at time slot t. If δ f (t) = 1, the UAV is in flight at time slot t, otherwise, δ f (t) = 0, P0, P0′ represent the constants of blade profile power and induced power of the UAV in hovering state, U tip is the tip speed of the rotor blade, v0 is the average rotor induced speed in the hovering state, and v t is the flight speed of the drone in time slot t, ξ d and r are the fuselage drag ratio and rotor reliability, ρ and S are the air density and rotor disc area, respectively; let E h (t) represents the hovering energy consumption of the UAV in time slot t, which is modeled as: E h (t) = δ h (t)(P0+P0′)τ, where δ h (t)∈{0,1} represents the hovering variable of the drone in time slot t. If δ h (t) = 1, the UAV is in hovering state at time slot t, otherwise, δ h (t) = 0; let E c (t), E s (t) represent the communication energy consumption and perception energy consumption of the UAV, which can be modeled as E c (t) = τTr(W c (t)(W c (t)) H ), 9. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 8, characterized in that: In step S8, the system cost function is modeled, specifically including: taking into account the user queue length and the drone perception performance, the system cost function is modeled as a weighted function of the longest user queue length and the Cramer-Rao bound of the maximum target angle estimation. Let η(t) represent the system cost function of time slot t, which can be modeled as Among them, β1 and β2 represent weight factors respectively.
10. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 9, characterized in that: In step S9, the communication perception constraint and the drone trajectory constraint are modeled, specifically including: any user must communicate at least once within the system time. This constraint can be expressed as: In any time slot, the drone can communicate with at most N users simultaneously. This constraint can be expressed as: Any target is sensed at least once within the system time. This constraint can be expressed as: In any time slot, the drone can sense at most one target. This constraint can be expressed as: The user transmission rate should meet the minimum communication rate threshold. This constraint can be expressed as: in represents the minimum communication rate threshold of the jth user; the Cramer-Rao bound of the drone's estimated target should be less than a threshold to meet the perception performance. This constraint can be expressed as: where γ th (t) represents the Cramer-Rao bound threshold of the drone’s estimated target; the total power perceived by the drone’s communication does not exceed its maximum transmit power. This constraint can be expressed as: Where P max represents the maximum transmission power of the UAV; the total energy consumption of the UAV should not exceed the total energy. This constraint can be expressed as: Where E max represents the total energy of the drone; the flight distance of the drone between two consecutive time slots should satisfy the distance constraint, which can be expressed as:
11. The method for optimizing flight trajectory and allocating resources of the unmanned aerial vehicle synaesthesia integrated system according to claim 10, characterized in that: In step S10, the flight trajectory of the UAV and the resource allocation strategy are determined based on the system cost function optimization, specifically including: the modeling system performance metric is: The UAV trajectory and resource allocation problem is modeled as a Markov decision process MDP problem, modeling the system state, action and benefit function; the system state in time slot t is in, represents the length of the user queue in time slot t, E r (t) represents the remaining energy of the UAV in time slot t, which can be expressed as: Let a(t) represent the action space of the drone in time slot t, which can be expressed as: Where z represents the flight distance of the drone in a time slot, z = v max τ; let r(t) represent the reward function of time slot t, and model it as follows: r(t) = -η(t)-h1+h2+h3, where h1 represents the penalty factor for the exhaustion of the drone battery energy, h2 represents the reward factor for all users to communicate at least once within the system time, and h3 represents the reward factor for all targets to be perceived at least once within the system time. The flight trajectory and resource allocation strategy of the drone are optimized based on the dual Q network DDQN framework and the greedy algorithm. Specifically, given the state and action of the MDP, the communication perception scheduling strategy of the drone is determined based on the greedy algorithm. Determine the UAV communication and perception precoding scheme W based on convex optimization method c (t), W s (t); UAV communication perception scheduling strategy based on given time slot And the precoding scheme W c (t), W s (t), Determine the flight trajectory of the drone based on long-term reward optimization
Citation Information
Cited By
Unmanned aerial vehicle communication perception integrated resource optimization method in emergency communication scene
CN121240057A
Multi-moving-target tracking method and system of communication and sensing integrated unmanned aerial vehicle
CN122218684A
A multi-moving target tracking method and system for integrated unmanned aerial vehicle
CN122218684B
