Method for resource allocation and trajectory optimization of intelligent reflecting surface assisted unmanned aerial vehicle cognitive network

Abstract

This invention discloses a resource allocation and trajectory optimization method for a UAV cognitive network based on a smart reflector: Step 1: Construct an air-to-ground communication model and solve for the channel gain; Step 2: Obtain the signal-to-noise ratio expression for the secondary UAV user in the nth time slot and the interference expression caused by the cognitive base station at the Kth primary ground user using the channel gain; Step 3: Based on the signal-to-noise ratio expression for the secondary UAV user, derive the average reachable rate expression for the secondary UAV user within N time slots; Step 4: Under the constraints of interference temperature of the primary ground user, reflection coefficient of the smart reflector, UAV maneuverability, and power limitations of the cognitive base station, jointly optimize the beamforming vector of the cognitive base station, the flight trajectory of the UAV, and the phase shift matrix of the smart reflector to establish a framework for maximizing the average reachable rate of the secondary UAV user; Step 5: Solve the optimization problem using an alternating optimization method.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a resource allocation and trajectory optimization method based on a smart reflective surface-assisted UAV cognitive network. Background Technology

[0002] In recent years, with continuous technological advancements and cost reductions, drones have become increasingly widely used in freight transport, aerial photography, surveillance, search and rescue, and other fields. Simultaneously, the rapid development of 5G and B5G networks is focusing more and more services on applications in uninhabited areas. Compared to terrestrial communication, drone communication has the following advantages: Drones, flying in the air, can establish strong air-to-ground line-of-sight links, providing better channel conditions. It's even possible to predict the channel state information and corresponding communication performance of drones at different three-dimensional positions based on the location information of ground nodes; drones have flexible deployment capabilities, providing communication services in complex terrain and areas difficult to cover by traditional terrestrial communication base stations; the high maneuverability of drones allows them to adjust their altitude and horizontal position while effectively avoiding ground interference sources, improving communication quality.

[0003] However, drone communication faces the challenge of limited spectrum resources. The licensed spectrum used for drone communication is finite, and it's difficult to allocate spectrum for new drones. To address this challenge, cognitive radio technology has been introduced, considered one of the effective methods to improve spectrum utilization efficiency. As an intelligent radio communication technology, cognitive radio can sense the surrounding radio spectrum environment and adaptively adjust according to the actual situation, aiming to improve spectrum utilization by effectively using radio spectrum resources to meet the ever-increasing demand for wireless communication. By identifying idle spectrum resources in real-time through radio spectrum environment sensing, drones can temporarily use these resources, thereby improving spectrum utilization efficiency and alleviating the strain on spectrum resources for drone communication. Furthermore, this network can establish flexible communication links according to actual needs, enabling drones to communicate more effectively with other drones or ground base stations, reducing communication interference and improving communication quality.

[0004] Intelligent reflectors are a technology that proactively reconfigures the radio propagation environment by controlling multiple small reflective elements to alter signal properties. In UAV cognitive networks, intelligent reflectors are introduced to address issues such as conflicting performance enhancements between primary and secondary users, and line-of-sight link congestion caused by complex terrain. By optimizing the phase offset matrix of the reflector, the received signal energy at the target user can be enhanced, while the signal strength at non-target users can be suppressed, thereby improving system performance and reducing interference. Simultaneously, intelligent reflectors can also use directional reflection to bypass obstacles, reducing channel quality attenuation and improving communication performance. Therefore, in UAV cognitive networks, intelligent reflector technology can help resolve performance conflicts between primary and secondary users and line-of-sight link congestion caused by complex terrain. Regarding performance conflicts between primary and secondary users, optimizing the phase offset matrix of the intelligent reflector enhances the received signal energy at the target user and suppresses the received signal strength at non-target users, thereby improving system performance while reducing interference. As for line-of-sight link congestion caused by complex terrain, intelligent reflectors use directional reflection of incident signals to bypass obstacles. Therefore, line-of-sight transmission can be maintained even in complex environments, thereby reducing channel quality degradation and improving communication performance. Thus, smart reflectors play a crucial role in improving the communication performance of UAV-assisted cognitive radio networks. However, in existing technologies, in scenarios where smart reflectors assist UAV cognitive network communication, secondary UAV users cannot achieve optimal transmission rates in the presence of obstacles. Summary of the Invention

[0005] To address the aforementioned problems in existing technologies, this invention significantly improves system performance and increases the rate of secondary drone users by deploying intelligent reflective surfaces. It provides a resource allocation and trajectory optimization method based on intelligent reflective surface-assisted drone cognitive networks, jointly optimizing the intelligent reflective surface phase shift matrix, drone flight trajectory, and transmission power to maximize the average rate of secondary drone users.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution:

[0007] A resource allocation and trajectory optimization method based on a smart reflector-assisted UAV cognitive network is proposed. The wireless communication system operated by this method includes a cognitive base station with L antennas, a secondary UAV user, and K primary ground users, where the user set is represented as... The method includes the following steps:

[0008] Step 1: Establish an air-to-ground communication model: Considering that there may be many obstacles in the surrounding area during UAV cognitive communication, causing line-of-sight links to be blocked, multiple Loss links can be created to enhance communication by deploying intelligent reflectors. In this case, the channels between the cognitive base station and the secondary UAV user, the cognitive base station and the kth primary user on the ground, the intelligent reflector and the secondary UAV user, and the intelligent reflector and the kth primary user on the ground are modeled to form an air-to-ground communication model, and its channel gain is solved.

[0009] Step 2: Using the channel gain from Step 1, obtain the signal-to-noise ratio expression for the secondary UAV user in the nth time slot, and the interference expression caused by the cognitive base station at the kth major ground user, and proceed with the next step of analysis;

[0010] Step 3: Based on the signal-to-noise ratio expression for secondary drone users in Step 2, derive the expression for the average reachable rate of the system for secondary drone users within each time slot;

[0011] Step 4: Under the constraints of interference temperature of the main ground users, reflection coefficient of the smart reflector, UAV maneuverability and cognitive base station power limit, jointly optimize the beamforming vector of the cognitive base station, the flight trajectory of the UAV and the phase shift matrix of the smart reflector to establish a framework for maximizing the average reachability of secondary UAV users.

[0012] Step 5: Use an alternating optimization method to solve the optimization problem established in Step 4, and optimize the base station's transmit beamforming, the intelligent reflector's reflection phase shift, and the UAV's flight trajectory.

[0013] Preferably, step 5 specifically involves: first, fixing the reflection phase shift of the smart reflector and the UAV flight trajectory, and optimizing the base station's transmit beamforming; second, fixing the base station's transmit beamforming and the UAV flight trajectory, and optimizing the reflection phase shift of the smart reflector; and finally, fixing the reflection phase shift of the smart reflector and the base station's transmit beamforming, and optimizing the UAV flight trajectory.

[0014] Furthermore, the distribution of each communication node in step 1 is defined as follows:

[0015] All communication nodes are placed in a three-dimensional Cartesian coordinate system, and an intelligent reflective surface with M reflective units is deployed on the building surface. The coordinates of the ground-based cognitive base station are... The coordinates of the kth primary ground user are T is divided into N smaller time slots, each time slot having a length of... The horizontal coordinate of the secondary drone user in the nth time slot is Suppose a smart reflector with M reflective elements is used as a uniform linear array. In the nth time slot, the phase offset matrix of the smart reflector is represented as a diagonal matrix:

[0016]

[0017] in, .

[0018] The model from the cognitive base station to the secondary drone user is a Rayleigh channel model, with the channel gain being:

[0019]

[0020] Where C0 represents the channel gain at a reference distance d0 = 1 m, This is the path loss index. Let be the distance between the cognitive base station and the secondary drone user in the nth time slot. It is a complex Gaussian random variable with zero mean and unit variance.

[0021] The channel gain from the cognitive base station to the k-th primary user on the ground is:

[0022]

[0023] in, It is the path loss coefficient. Let be the distance between the cognitive base station and the k-th primary user. It is a complex Gaussian random variable with zero mean and unit variance.

[0024] The channel gain between the smart reflector and the secondary drone user is:

[0025]

[0026] in, The distance between the secondary drone user and the smart reflector in the nth time slot. These represent the antenna spacing and carrier wavelength, respectively. This represents the cosine of the arrival angle of the intelligent reflector in each time slot. u This is the altitude of secondary drone users, w r It is the horizontal position of the intelligent reflective surface.

[0027] The gain of the smart reflector to the k-th primary user channel on the ground is expressed as:

[0028]

[0029] The channel gain from the cognitive base station to the intelligent reflector can be written as:

[0030]

[0031] in, The path loss index, It is the Rice factor. Let be the distance between the smart reflective surface and the k-th primary user on the ground. To determine the distance between the cognitive base station and the intelligent reflective surface. , Let cosine be the angle of deviation from the smart reflective surface to the k-th primary user. This represents the cosine of the angle of arrival from the cognitive base station to the intelligent reflector. All are independent complex Gaussian distributions with zero mean and unit variance.

[0032] Furthermore, in step 2, the signal-to-noise ratio calculation formula for the secondary drone user in the nth time slot is:

[0033]

[0034] Where f[n] is the transmit beamforming vector of the cognitive base station in the nth time slot. This is the power of additive Gaussian noise at the secondary drone user location.

[0035] The interference caused by the cognitive base station at the k-th major ground user is represented as:

[0036] .

[0037] Furthermore, in step 3, the expression for the average reachability of the system for secondary drone users within N time slots is:

[0038]

[0039] Furthermore, in step 4, the problem P0 for maximizing the average reachability rate of secondary drone users is established as follows:

[0040]

[0041] Wherein, Equation C1 represents the transmit power constraint of the cognitive base station, with a maximum power limit of P. max Formula C2 represents the phase adjustment constraint of the intelligent reflector; Formula C3 represents the interference from the cognitive base station's signal to the primary ground user receiver, where φ is the interference threshold for the primary ground user, i.e., the maximum interference limit; Formulas C4 and C5 represent the position and maneuverability constraints for secondary UAV users, respectively; V max That is the maximum flight speed of the drone. These are the starting and ending points of the drone, respectively; P maxThe maximum power limit for the transmission power of the cognitive base station; 𝛤 is the interference threshold for the main ground users, i.e., the maximum limit of interference.

[0042] Furthermore, step 5 employs an alternating optimization method to solve the established optimization problem, specifically including the following steps:

[0043] Step 5-1: Treating the flight trajectory Q of the secondary drone user and the phase shift matrix φ of the smart reflector as constants, the beamforming vector of the cognitive base station is optimized. Therefore, the optimization problem P0 is reformulated as follows:

[0044]

[0045] C1 represents the transmit power limit of the cognitive base station, and C2 represents the interference limit of the cognitive base station to major ground users.

[0046] Step 5-2: Define the following matrix:

[0047]

[0048] Correspondingly, in the non-concave objective function of P1, R s [n] can be rewritten as:

[0049]

[0050] Meanwhile, the left side of the inequality in constraint C2 in P1 can be restated as:

[0051]

[0052] Define matrix for:

[0053]

[0054] Therefore, the optimization problem P0 can be rewritten as:

[0055]

[0056] Where C1 and C2 are the definition matrices. The constraints that come with this are: C3 is the transmit power limit of the cognitive base station; C4 is the interference limit of the cognitive base station to the main users on the ground.

[0057] Step 5-3: Using the semidefinite relaxation technique, relax the rank-1 constraint C2 in P2, transforming the problem into a convex problem:

[0058]

[0059] The optimization problem P3 is convex. It can be solved efficiently using MATLAB toolboxes (such as CVX).

[0060] Step 5-4: Given a fixed flight trajectory Q of the secondary drone user and beamforming f of the cognitive base station, optimize the phase shift matrix φ of the intelligent reflector. The optimization problem can then be written as:

[0061]

[0062] Step 5-5: Introduction And define:

[0063]

[0064] R in the objective function of P4 s [n] is written as:

[0065]

[0066] The left-hand side of the inequality in P4, C2, is restated as follows:

[0067]

[0068] Introduction It can be written as:

[0069]

[0070] Furthermore, by utilizing semidefinite relaxation, the optimization problem can be reformulated as:

[0071]

[0072] Steps 5-6: Fix the phase shift matrix φ of the intelligent reflector and the beamforming f of the cognitive base station, and optimize the flight trajectory Q of the secondary UAV user. The optimization problem can be reformulated as:

[0073]

[0074] Rewrite the non-concave objective function 𝑅𝑠[𝑛] as:

[0075]

[0076] in .

[0077] Introducing auxiliary variables The objective function of P5 can be rewritten as:

[0078]

[0079] Problem P5 can be reformulated as:

[0080]

[0081] Steps 5-7: At the given point Perform a first-order Taylor expansion to obtain its lower bound. ,Right now

[0082]

[0083] in

[0084]

[0085] right Performing a first-order Taylor expansion, we obtain the following inequality:

[0086]

[0087] The optimization problem P7 is transformed into a convex problem, which can be restated as follows:

[0088]

[0089] The beneficial effects of this invention are:

[0090] When drones are secondary users, this invention proposes to apply intelligent reflective surfaces to the drone cognitive network. Under the constraints of interference temperature of the primary user, reflection coefficient of intelligent reflective surfaces, drone maneuverability and power limitation of cognitive base station, the beamforming vector of cognitive base station, flight trajectory of drone and phase shift matrix of intelligent reflective surface are jointly optimized to establish a framework for maximizing the average reachability of secondary drone users.

[0091] In this invention, since the problem is highly non-convex, to solve it, the block coordinate descent method is first used to divide the problem into three optimization sub-problems: (1) optimization of the phase shift matrix of the intelligent reflector; (2) optimization of the transmit power of the UAV base station; and (3) optimization of the flight trajectory of the UAV. The closed-form solutions of sub-problems (1) and (2) are obtained, and then substituted into sub-problem (3). An approximate solution to the problem is obtained by using the continuous convex approximation technique. Simulation results show that, compared with the existing benchmark scheme, the iterative optimization scheme proposed in this invention can improve the average reachability rate of secondary users.

[0092] Compared with existing UAV communication technologies that do not deploy intelligent reflectors, this invention utilizes intelligent reflectors to maintain line-of-sight links, improves the average reachability rate for secondary UAV users and reduces interference to primary users; it also reduces the power consumption of the entire system and improves the flexibility of the entire communication system. Attached Figure Description

[0093] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments, and the advantages of the present invention in the above and / or other aspects will become clearer.

[0094] Figure 1 This is a diagram of the cognitive UAV communication model based on intelligent reflective surface assistance of the present invention.

[0095] Figure 2 This is a drone trajectory diagram of a preferred embodiment of the present invention.

[0096] Figure 3 This is a graph showing the relationship between the number of reflective surfaces and the average reachability of secondary drone users in a preferred embodiment of the present invention.

[0097] Figure 4 This is a graph showing the relationship between the transmit power and the average reachability of secondary drone users in a preferred embodiment of the present invention.

[0098] Figure 5 This is a graph showing the relationship between the number of antennas and the average reachable rate of secondary drone users in a preferred embodiment of the present invention. Detailed Implementation

[0099] This invention relates to resource allocation and trajectory optimization in a cognitive UAV network assisted by intelligent reflectors. Considering the scenario where UAVs, as secondary users, share spectrum with primary ground users, intelligent reflectors are deployed on building surfaces to assist communication. To improve the average reachability of secondary users, this invention proposes a method for jointly optimizing the transmit beamforming vector of the cognitive base station, the phase offset matrix of the intelligent reflector, and the flight trajectory of the secondary UAV users. To address the non-convex problem, a block coordinate descent method is used to decompose the problem into three non-convex sub-problems, and semi-definite relaxation and continuous convex approximation techniques are employed to solve these sub-problems. This invention relates to an iterative optimization algorithm.

[0100] This embodiment's joint optimization method is based on a smart reflector-assisted UAV cognitive network. The communication system includes a cognitive base station with L antennas, a secondary UAV user, and K primary ground users, where the user set is represented as follows: .

[0101] The specific steps of the joint optimization method in this embodiment are as follows:

[0102] Step 1: Model the channels between the cognitive base station and the secondary UAV user, between the cognitive base station and the kth primary user on the ground, between the intelligent reflector and the secondary UAV user, and between the intelligent reflector and the kth primary user on the ground to form an air-to-ground communication model and solve for the channel gain.

[0103] Specifically, such as Figure 1As shown, all communication nodes are placed in a three-dimensional Cartesian coordinate system, and a smart reflective surface with M reflective units is deployed on the building surface. The coordinates of the ground-based cognitive base station are... The coordinates of the kth primary ground user are T is divided into N smaller time slots, each time slot having a length of... The horizontal coordinate of the secondary drone user in the nth time slot is Suppose a smart reflector with M reflective elements is used as a uniform linear array. In the nth time slot, the phase offset matrix of the smart reflector is represented as a diagonal matrix. .

[0104] The model from the cognitive base station to the secondary drone user is a Rayleigh channel model, with the channel gain being:

[0105]

[0106] Where C0 represents the channel gain at a reference distance d0 = 1 m, This is the path loss index. Let be the distance between the cognitive base station and the secondary drone user in the nth time slot. It is a complex Gaussian random variable with zero mean and unit variance.

[0107] The channel gain from the cognitive base station to the k-th primary user on the ground is:

[0108]

[0109] in, It is the path loss coefficient. Let be the distance between the cognitive base station and the k-th primary user. It is a complex Gaussian random variable with zero mean and unit variance.

[0110] The channel gain between the smart reflector and the secondary drone user is:

[0111]

[0112] in, The distance between the secondary drone user and the smart reflector in the nth time slot. These represent the antenna spacing and carrier wavelength, respectively. This represents the cosine of the arrival angle of the intelligent reflector in each time slot. u This is the altitude of secondary drone users, w r It is the horizontal position of the intelligent reflective surface.

[0113] The gain of the smart reflector to the k-th primary user channel on the ground is expressed as:

[0114]

[0115] The channel gain from the cognitive base station to the intelligent reflector can be written as:

[0116]

[0117] in, The path loss index, It is the Rice factor. Let be the distance between the smart reflective surface and the k-th primary user on the ground. To determine the distance between the cognitive base station and the intelligent reflective surface. , Let cosine be the angle of deviation from the smart reflective surface to the k-th primary user. This represents the cosine of the angle of arrival from the cognitive base station to the intelligent reflector. All are independent complex Gaussian distributions with zero mean and unit variance.

[0118] Step 2 uses the channel gain from Step 1 to obtain the signal-to-noise ratio expression for the secondary UAV user in the nth time slot, and the interference expression caused by the cognitive base station at the Kth major ground user, for further analysis.

[0119] Specifically, the signal-to-noise ratio calculation formula for secondary drone users in the nth time slot can be written as:

[0120]

[0121] Where f[n] is the transmit beamforming vector of the cognitive base station in the nth time slot. This is the power of additive Gaussian noise at the secondary drone user location.

[0122] The interference caused by the cognitive base station at the k-th major ground user is represented as:

[0123]

[0124] Step 3: Based on the signal-to-noise ratio expression for the secondary drone user in Step 2, derive the expression for the average achievable rate of the system for the secondary drone user within N time slots.

[0125] Specifically, the average reachability and rate of the system for secondary drone users over N time slots are expressed as follows:

[0126]

[0127] Step 4: Under the constraints of interference temperature for primary ground users, reflection coefficient of intelligent reflectors, UAV maneuverability, and cognitive base station power limitations, jointly optimize the flight trajectories of secondary UAV users. Phase shift matrix of intelligent reflective surface Beamforming vectors of cognitive base stations Establish a framework for maximizing the average reachability of secondary drone users.

[0128] Specifically, the problem P0 for maximizing the average reachability of secondary drone users is:

[0129]

[0130] Wherein, Equation C1 represents the transmit power constraint of the cognitive base station, with a maximum power limit of P. max Formula C2 represents the phase adjustment constraint of the intelligent reflector; Formula C3 represents the interference from the cognitive base station's signal to the primary ground user receiver, where φ is the interference threshold for the primary ground user, i.e., the maximum interference limit; Formulas C4 and C5 represent the position and maneuverability constraints for secondary UAV users, respectively, and V... max That is the maximum flight speed of the drone. These are the starting point and the ending point of the drone, respectively.

[0131] Step 5: Solve the optimization problem established in Step 4 using an alternating optimization method, optimizing the base station's transmit beamforming, the intelligent reflector's reflection phase shift, and the UAV's flight trajectory. Specifically, this includes the following steps:

[0132] Step 5-1: Flight trajectory Q of the secondary UAV user and phase shift matrix of the smart reflector Treating it as a constant, the beamforming vector f of the cognitive base station is optimized. Therefore, the optimization problem P0 is reformulated as follows:

[0133]

[0134] Step 5-2: Define the following matrix:

[0135]

[0136] Correspondingly, in the non-concave objective function of P1, R s [n] can be rewritten as:

[0137]

[0138] Meanwhile, the left side of the inequality in constraint C2 in P1 can be restated as:

[0139]

[0140] Define matrix for:

[0141]

[0142] Therefore, optimization problem P1 can be rewritten as:

[0143]

[0144] Step 5-3: Using the positive semidefinite relaxation technique, relax the rank-one constraint, transforming the problem into a convex problem:

[0145]

[0146] The optimization problem P3 is convex. It can be solved efficiently using MATLAB toolboxes (such as CVX).

[0147] Step 5-4: Fix the flight trajectory Q of the secondary drone user and the beamforming f of the cognitive base station, and apply the phase shift matrix to the intelligent reflector. If we perform optimization, the optimization problem can be written as:

[0148]

[0149] Step 5-5: Introduction And define:

[0150]

[0151] R in the objective function of P4 s [n] is written as:

[0152]

[0153] The left-hand side of the inequality in P4, C2, is restated as follows:

[0154]

[0155] Introduction It can be written as:

[0156]

[0157] Furthermore, by utilizing semidefinite relaxation, the optimization problem can be reformulated as:

[0158]

[0159] Steps 5-6: Fix the phase shift matrix of the smart reflector The beamforming f of the cognitive base station optimizes the flight trajectory Q of secondary drone users. The optimization problem can be reformulated as:

[0160]

[0161] Rewrite the non-concave objective function 𝑅𝑠[𝑛] as:

[0162]

[0163] in .

[0164] Introducing auxiliary variables The objective function of P5 can be rewritten as:

[0165]

[0166] Problem P5 can be reformulated as:

[0167]

[0168] Steps 5-7: At the given point Perform a first-order Taylor expansion to obtain its lower bound. ,Right now

[0169]

[0170] in

[0171]

[0172] right Performing a first-order Taylor expansion, we obtain the following inequality:

[0173]

[0174] The optimization problem P7 is transformed into a convex problem, which can be restated as follows:

[0175] .

[0176] This invention uses the block coordinate descent method to decompose the original problem into three subproblems. It uses semidefinite relaxation and continuous convex approximation techniques to transform the non-convex problem into convex problems P3, P5, and P8. Then, it sequentially solves the secondary UAV user trajectory, the intelligent reflector phase shift matrix, and the cognitive base station transmit beamforming vector, and uses alternating optimization iteration until convergence.

[0177] Based on the above embodiments, data simulation is performed: The diagrams and specific parameter values ​​provided in the following embodiments are mainly for illustrating the basic concept of the present invention and for simulating and verifying the invention. In specific application environments, appropriate adjustments can be made according to the actual scenario and requirements.

[0178] Assume the initial horizontal position of the secondary drone user in the communication system is q0 = [-500, 20] m, and the final horizontal position of the secondary drone user is q f =[500, 20] m, the altitude z of the secondary drone user u = 80 m, maximum speed V of secondary drone users max = 20 m / s; the horizontal positions of ground main user 1 and ground main user 2 are w1 = [−200, 10] m and w2 = [200, 10] m respectively; the horizontal position of the smart reflector is w r = [0, 0] m, height is 40 m; the horizontal position of the Cognitive Base Station (CBS) is w b = [0, 40] m; Maximum transmission power P of the cognitive base station max 1 w; time slot length =1 s; Number of reflecting units M = 20; Noise power = -80dBm, interference threshold = -70 dBm; Path loss index Path loss at reference distance Iteration accuracy The antenna spacing is .

[0179] The present invention is a scheme with RIS, and the following benchmark schemes are set: initial trajectory scheme: a scheme that jointly optimizes the phase shift matrix of the intelligent reflector and the beamforming vector transmitted by the cognitive base station (without optimizing the trajectory of secondary UAV users); w / oRIS scheme: a scheme that jointly optimizes the trajectory of secondary UAV users and the beamforming vector transmitted by the cognitive base station (without intelligent reflector).

[0180] Figure 2The flight trajectories under different schemes are shown when the interference threshold is −30 dBm. It can be observed that in the w / o RIS scheme, secondary drone users fly in a straight line towards the cognitive base station, hover above it, and eventually return to their starting position. This is because there is no intelligent reflector in this scheme; secondary users can only communicate with the cognitive base station via a direct link. To maximize their average reachability and speed, secondary drone users will try to get as close to the cognitive base station as possible. In contrast, the flight trajectory using the scheme of this invention is significantly different from the w / o RIS scheme. This is because the joint optimization algorithm proposed in this invention can optimize the drone trajectory, the phase of the reflector, and the base station's transmitted beamforming vector to balance the gain for secondary drone users and the interference for primary users. Therefore, in the scheme of this invention, the flight trajectory of secondary drone users is more flexible, not only utilizing the gain of the reflector but also better avoiding interference to primary users.

[0181] Figure 3 The system performance under different numbers of intelligent reflector units (r) is demonstrated. When comparing the three schemes, the scheme with RIS improves system performance compared to the other two. Specifically, compared to the initial trajectory scheme, the scheme of this invention optimizes the UAV trajectory, allowing secondary UAV users to approach the cognitive base station, reducing channel fading, enhancing signal quality, and thus increasing the average reachability rate of secondary users, thereby improving the system's communication quality. Simultaneously, compared to the w / o RIS scheme, the scheme with RIS introduces intelligent reflectors, enabling better superposition of reflected and transmitted signals. By optimizing the intelligent reflector phase and the base station's transmit beamforming vector, signal enhancement is achieved for secondary UAV users, while interference is reduced for primary users, thus improving system performance. This effect becomes more pronounced with increasing numbers of intelligent reflector units. Simulation results of this invention demonstrate that the scheme of this invention can significantly improve system communication performance under different r values ​​by jointly optimizing the intelligent reflector phase, transmit beamforming vector, and UAV trajectory.

[0182] Figure 4The average reachability of secondary UAV users under different maximum transmit powers is presented. First, it can be seen that under the scheme of this invention, the average reachability gradually increases with the increase of the maximum transmit power. This is because the signal quality improves with the increase of transmit power, thus allowing the transmission of more information. Second, compared with the scheme of this invention, the average reachability of the w / o RIS scheme is lower at all transmit power levels, and the gap gradually widens with the increase of transmit power. This is because the w / o RIS scheme does not introduce a smart reflector, thus failing to enhance the quality of the reflected signal, and is also affected by factors such as channel fading, resulting in relatively low signal quality. In contrast, although the initial trajectory scheme considers the phase of the smart reflector and the transmit power of the secondary base station, it does not optimize the trajectory of the secondary UAV users, resulting in relatively weaker performance. In summary, the scheme of this invention is superior to both the w / o RIS scheme and the initial trajectory scheme; therefore, this invention can be considered a feasible solution to improve the performance of communication systems.

[0183] from Figure 5 As can be seen, under all cognitive base station antenna counts, the proposed solution achieves better performance than both the initial trajectory solution and the w / o RIS solution. Specifically, compared to the initial trajectory solution, the proposed solution optimizes the UAV trajectory and reflector phase, and simultaneously optimizes the transmit beamforming vector of the secondary base station, thereby better balancing interference between primary and secondary users and improving the system's average reachability. In contrast, the initial trajectory solution only optimizes the UAV trajectory and the transmit beamforming vector of the secondary base station without introducing intelligent reflector optimization, resulting in weaker performance. Similarly, the w / o RIS solution only optimizes the transmit beamforming vector of the secondary base station and the trajectory of secondary UAV users, without considering the performance improvement effect of the reflector. Therefore, the w / o RIS solution also performs weaker than the proposed solution. Simulation results show that the proposed solution exhibits superior performance under different transmit power and antenna count levels.

[0184] This invention investigates the resource allocation and trajectory optimization problems of UAV cognitive base stations assisted by intelligent reflectors. Specifically, it considers scenarios where UAVs act as aerial base stations assisting cognitive radio networks, taking into account the interruption of line-of-sight (LAS) transmission between the UAV base station and ground users in complex terrain, and reconstructs LAS transmission using intelligent reflectors. Under constraints imposed by primary user interference temperature, intelligent reflector reflection coefficient, and UAV maneuverability, the average reachability rate for secondary users is maximized through the joint design of the intelligent reflector phase shift matrix, UAV flight trajectory, and transmission power.

[0185] This invention can effectively alleviate the problem of spectrum resource scarcity, provide high flexibility for resource allocation in cognitive UAV networks through intelligent reflective surfaces, and suppress interference and enhance useful signals without additional energy consumption; it can reconstruct obstructed line-of-sight transmission for cognitive UAV networks, expand the service range of cognitive UAV networks, and enhance communication.

[0186] It should be noted that the above embodiments are merely illustrative of implementation methods of the present invention, and are not intended to limit the present invention. The present invention is not limited to the examples described above. Those skilled in the art can make improvements and modifications without departing from the spirit of the present invention, and these modifications all fall within the scope of protection of the present invention.

Claims

1. A resource allocation and trajectory optimization method based on a smart reflector-assisted UAV cognitive network, wherein the network includes a cognitive base station with L antennas, a secondary UAV user, and K primary ground users, wherein the set of primary ground users is represented as... Its characteristics are: The method includes the following steps: Step 1: Model the channels between the cognitive base station and the secondary UAV user, the cognitive base station and the kth primary ground user, the intelligent reflector and the secondary UAV user, and the intelligent reflector and the kth primary ground user to form an air-to-ground communication model and solve for the channel gain. Step 2: Using the channel gain from Step 1, obtain the signal-to-noise ratio expression for the secondary UAV user in the nth time slot, and the interference expression caused by the cognitive base station at the Kth primary ground user. Step 3: Based on the signal-to-noise ratio expression for secondary drone users in Step 2, derive the expression for the average achievable rate of the communication system for secondary drone users within N time slots; Step 4: Under the constraints of interference temperature of the main ground users, reflection coefficient of the smart reflector, UAV maneuverability and cognitive base station power limit, jointly optimize the beamforming vector of the cognitive base station, the flight trajectory of the UAV and the phase shift matrix of the smart reflector to establish a framework for maximizing the average reachability of secondary UAV users. Step 5: Use an alternating optimization method to solve the optimization problem established in Step 4, and optimize the base station's transmit beamforming, the intelligent reflector's reflection phase shift, and the UAV's flight trajectory. In step 1, the intelligent reflective surface, the distribution of K primary ground users, and the status of secondary UAV users are defined as follows: All communication nodes are placed in a three-dimensional Cartesian coordinate system, and an intelligent reflective surface with M reflective units is deployed on the building surface; the coordinates of the ground cognitive base station are... The coordinates of the kth primary ground user are T is divided into N time slots, each time slot having a length of... The horizontal coordinate of the secondary drone user in the nth time slot is Suppose a RIS with M reflective elements is used as a uniform linear array. In the nth time slot, the phase offset matrix of the smart reflector is represented as a diagonal matrix: in, ; The model from the cognitive base station to the secondary drone user is a Rayleigh channel model, with the channel gain being: Where C0 represents the channel gain at a reference distance d0 = 1 m, α The path loss index; The distance between the cognitive base station and the secondary drone user in the nth time slot; It is a complex Gaussian random variable with zero mean and unit variance; The channel gain from the cognitive base station to the k-th primary ground user is: in, It is the path loss coefficient. The distance between the cognitive base station and the kth primary ground user; It is a complex Gaussian random variable with zero mean and unit variance; The channel gain between the smart reflector and the secondary drone user is: in, The distance between the secondary drone user and the smart reflector in the nth time slot. These represent the antenna spacing and carrier wavelength, respectively. This represents the cosine of the RIS arrival angle in each time slot; z u This is the altitude of secondary drone users, w r It refers to the horizontal position of the intelligent reflective surface; The channel gain from RIS to the k-th primary ground user is expressed as: The channel gain from the cognitive base station to the RIS is expressed as: in, This is the path loss index. Rice factor; Let be the distance between RIS and the k-th primary ground user. To determine the distance between the cognitive base station and the RIS; , Let cosine be the angle of deviation from the RIS to the k-th primary ground user. This represents the cosine of the angle of arrival from the cognitive base station to the RIS. All are independent complex Gaussian distributions with zero mean and unit variance.

2. The method as described in claim 1, characterized in that: In step 2, the signal-to-noise ratio calculation formula for secondary drone users in the nth time slot is as follows: Where f[n] is the transmit beamforming vector of the cognitive base station in the nth time slot. It is the power of additive Gaussian noise at the secondary drone user location; The interference caused by the cognitive base station at the k-th major ground user is represented as: 。 3. The method as described in claim 2, characterized in that: In step 3, the expression for the average reachability of the communication system for secondary drone users within N time slots is: 。 4. The method as described in claim 3, characterized in that: In step 4, the flight trajectories of secondary drone users are jointly optimized. Phase shift matrix of intelligent reflective surface Beamforming vectors of cognitive base stations The problem of maximizing the average reachability rate of secondary drone users, P0, is established as follows: Among them, V max The maximum flight speed of the drone is q0 and q f These are the starting and ending points of the drone, respectively; P max To limit the maximum power of the base station's transmission power; It is the interference threshold for the main ground users, i.e., the maximum extent to which they are subject to interference.

5. The method according to any one of claims 1-4, characterized in that: Step 5 is as follows: Fix the reflection phase shift of the smart reflector and the UAV flight trajectory, and optimize the transmission beamforming of the base station; fix the transmission beamforming of the base station and the UAV flight trajectory, and optimize the reflection phase shift of the smart reflector; fix the reflection phase shift of the smart reflector and the transmission beamforming of the base station, and optimize the UAV flight trajectory.

6. The method as described in claim 4, characterized in that: Step 5 employs an alternating optimization method to solve the established optimization problem, specifically including the following steps: Step 5-1: Flight trajectory Q of the secondary UAV user and phase shift matrix of the smart reflector Treating it as a constant, the beamforming vector f of the cognitive base station is optimized. Therefore, the optimization problem P0 is reformulated as follows: Step 5-2: Define the following matrix: Therefore, the optimization problem P1 is rewritten as follows: Step 5-3: Using the positive semidefinite relaxation technique, relax the rank-one constraint. The problem then becomes: The optimization problem P3 is convex; Step 5-4: Fix the flight trajectory Q of the secondary drone user and the beamforming f of the cognitive base station, and apply the phase shift matrix of RIS. If optimization is performed, the optimization problem can be represented as follows: Step 5-5: Introduction , And define: R in the objective function of P4 s [n] is written as: The left-hand side of the inequality in P4, C2, is restated as follows: Introduction Written as: After applying semidefinite relaxation, the optimization problem is reformulated as follows: Steps 5-6: Fix the phase shift matrix of the smart reflector The beamforming f of the cognitive base station is used to optimize the flight trajectory Q of secondary drone users. The optimization problem is reformulated as follows: In the non-concave objective function, 𝑅 𝑠 [𝑛] rewritten as: in ; Introducing auxiliary variables The objective function for P5 is rewritten as follows: Question P5 is rephrased as: Steps 5-7: At the given point Perform a first-order Taylor expansion to obtain its lower bound. ,Right now in right Performing a first-order Taylor expansion, we obtain the following inequality: The optimization problem P7 is transformed into a convex problem, which can be restated as follows: 。

Images