Hybrid optimization method and system for unmanned aerial vehicle multiple-input multiple-output communication network, device and medium
An optimization method for UAV communication systems that combines dynamic user grouping and NOMA technology solves the problems of resource allocation and interference in multi-user, multi-UAV environments, achieving efficient resource utilization and interference management, and improving system performance and communication quality.
Patent Information
- Application Number
- PCT/CN2025/129315
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-15
- Filing Date
- 2025-10-22
- Publication Date
- 2026-11-19
AI Technical Summary
Existing drone communication systems suffer from problems such as low resource allocation efficiency, insufficient interference management, and excessive system complexity in multi-user and multi-drone environments, especially in high-density scenarios where effective resource optimization is difficult to achieve.
A hybrid optimization framework based on cluster reconfiguration is adopted. By combining dynamic user grouping, LED selection, power allocation and NOMA technology, channel gain is calculated using the Lambertian radiation distribution. The optimization is then performed by combining the branch-and-bound method and the alternating direction multiplier method, so as to achieve efficient resource allocation and interference management.
It significantly improves spectrum efficiency and system throughput, reduces system complexity and energy consumption, and enhances communication reliability and stability, making it suitable for multi-cell UAV communication environments.
Smart Images

Figure CN2025129315_19112026_PF_FP_ABST
Abstract
Description
Hybrid optimization methods, systems, equipment, and media for UAV multi-input multi-output communication networks Technical Field
[0001] This invention relates to the field of unmanned aerial vehicle (UAV) communication and resource management optimization technology, specifically to a hybrid optimization method, system, device, and medium for UAV multi-input multi-output communication networks. Background Technology
[0002] In unmanned aerial vehicle (UAV) communication, visible light communication (VLC) has become a highly promising communication method due to its high bandwidth, low interference, and high security. However, UAV communication scenarios involve complex environments with multiple users and multiple cells, and traditional VLC systems face problems such as low resource allocation efficiency and difficulty in interference management. Therefore, how to achieve efficient resource allocation and interference management in UAV scenarios has become an urgent problem to be solved.
[0003] To address these challenges, massive MIMO technology has been introduced into UAV communication systems, significantly improving transmission efficiency and system capacity through multi-antenna configurations. However, in multi-user and multi-UAV environments, massive MIMO systems face unique challenges in both the uplink and downlink. In the downlink, multi-user and multi-UAV environments often lead to severe interference problems, especially at the edges of UAV coverage areas, where user equipment may receive overlapping signals from adjacent UAVs, resulting in cross-UAV interference and reduced data transmission efficiency. Existing research addresses downlink interference by grouping users, attempting to mitigate edge interference by assigning users with similar channel conditions to the same group. However, these methods often perform poorly when dealing with complex channel state changes in multi-UAV environments and lack dynamic adaptability, making it difficult to cope with frequent channel fluctuations in high-density multi-user scenarios. In the uplink, system capacity and resource allocation are the main challenges. As the number of users increases, the uplink capacity of a single UAV often cannot meet the transmission needs of multiple users, leading to low resource utilization. Furthermore, the large number of antennas in the system exacerbates these problems; activating all antennas simultaneously not only increases energy consumption but also significantly increases computational complexity. Existing research shows that efficient antenna selection can effectively reduce system complexity and energy consumption while optimizing uplink capacity. However, current research typically focuses on capacity enhancement in a single uplink scenario, failing to achieve comprehensive resource optimization in multi-user and multi-UAV environments.
[0004] CN118890636A proposes a joint optimization method for multi-UAV-assisted communication in wireless networks, aiming to improve communication reliability and fairness for ground users. This method constructs a joint optimization model with the objective of maximizing the minimum signal-to-noise ratio, considering constraints such as UAV and channel allocation, transmit power, and co-channel interference. It then uses clustering, genetic, and particle swarm optimization algorithms for iterative optimization, ultimately obtaining the optimal UAV-ground user matching, channel allocation, and transmit power allocation schemes. This method is suitable for cognitive wireless network environments with limited spectrum resources, especially in emergency situations where ground base stations fail due to natural disasters, enabling flexible deployment of UAV swarms to provide temporary communication services to ground users. Its advantages include improved communication reliability, spectrum utilization, and fairness for ground users; however, its disadvantages include high computational complexity, slow iterative convergence speed, and dependence on algorithm parameter settings, making it difficult to achieve ideal conditions.
[0005] CN116709384A proposes an optimization method for UAV NOMA (Non-Orthogonal Multiple Access) assisted communication based on a combination of convex optimization and machine learning, aiming to improve spectrum utilization, reduce latency, and enhance UAV energy efficiency. This method constructs a UAV-NOMA assisted communication model, optimizes UAV service allocation using the K-means clustering algorithm, and determines the transmit power and UAV trajectory for each user through a first-order Taylor expansion convex approximation. It is suitable for scenarios where UAV swarms provide task offloading for ground users, especially in computationally intensive and latency-sensitive applications such as virtual reality (VR) and simultaneous localization and mapping (SLAM). Its advantages include high spectrum utilization, short latency, high UAV energy efficiency, and the ability to improve transmission efficiency through multipath reception; its disadvantages include high computational complexity and dependence on the accuracy of environmental parameters and the training effect of the model.
[0006] CN115499805A proposes a joint optimization method for a multi-UAV data acquisition system based on non-orthogonal multiple access (NOMA), suitable for ground equipment data acquisition in emergency scenarios. This method uses multiple UAVs as mobile base stations to serve multiple ground devices in areas without ground base stations, utilizes NOMA technology for data transmission, and minimizes the system's task completion time by jointly optimizing UAV flight trajectories, ground equipment power control, and device grouping scheduling. The advantage of this method is its significantly improved task completion efficiency, making it particularly valuable in emergency data acquisition scenarios. However, this method is complex, computationally intensive, and may be affected by UAV energy limitations and changes in channel conditions in practical applications. Summary of the Invention
[0007] In view of the above-mentioned problems, the present invention is proposed.
[0008] Therefore, the technical problem solved by this invention is that existing technical solutions often suffer from low resource allocation efficiency, insufficient interference management, and excessive system complexity when dealing with resource allocation and interference issues in multi-UAV and multi-user environments. This invention proposes a dynamic resource allocation and interference management optimization method based on UAV swarms. It aims to improve the resource utilization and anti-interference capability of UAV communication systems, reduce system complexity, and achieve efficient real-time performance optimization by combining dynamic user grouping, LED selection, power allocation, and NOMA technology.
[0009] To address the aforementioned technical problems, this invention provides the following technical solution: a hybrid optimization method for multi-input multi-output communication networks of unmanned aerial vehicles (UAVs), comprising the following steps:
[0010] Based on a cluster reconstruction hybrid framework, a multi-cell, multi-UAV VLC system is deployed in the framework for scene setting and parameter configuration;
[0011] Each transmitter is configured to follow the Lambertian radiation distribution. User channel gain is calculated based on channel state information, and users are grouped.
[0012] The branch-and-bound method is used to optimize the selection of LED transmitters;
[0013] Based on the results of the user grouping, power is allocated using the user channel gain;
[0014] The alternating direction multiplier method is used to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system rate.
[0015] As a preferred embodiment of the hybrid optimization method for UAV multi-input multi-output communication networks described in this invention, the scenario setup includes: deploying a multi-cell, multi-UAV system, with each UAV equipped with multiple LED transmitters, and users randomly distributed and equipped with multiple PD receivers;
[0016] The parameter configuration includes initialization parameters for channel state information acquisition, MSE threshold, power allocation ratio, and LED selection range.
[0017] As a preferred embodiment of the hybrid optimization method for UAV multiple-input multiple-output communication networks described in this invention, the calculation of user channel gain includes:
[0018] Each LED is assumed to follow a Lambertian radiation distribution, and the user channel gain h is calculated based on channel state information. c,i,k,r :
[0019] Where m is the Lambertian exponent, m = -ln(2) / ln(cos(Φ)1 / 2 )), Φ 1 / 2 It is the half-power angle, A r It is the effective area of PD, d c,i,k,r φ is the distance between the i-th LED in the c-th group and the i-th PD of the k-th user. c,i,k,r and ψ c,i,k,r These are the incident angle and radiation angle of the i-th PD of the k-th user, respectively, T s (ψ c,i,k,r ) is the gain of the optical filter, g(ψ) c,i,k,r ) is the gain of the optical concentrator.
[0020] As a preferred embodiment of the hybrid optimization method for UAV multi-input multi-output communication networks described in this invention, the user grouping includes estimating the current channel of the user based on the user's channel state information, i.e., a set of characteristic parameters including channel gain and path loss between the user and the transmitter, and grouping users with similar estimation error levels into the same group.
[0021] User grouping is based on MSE, and the formula for calculating MSE is:
[0022] in, It is the estimated channel, MSE c,i,k,r To approximate the estimation error level, F is the Frobenius norm, and N... r To receive the number of PDs, N c,t This refers to the number of subcarriers.
[0023] Set the predefined MSE threshold ε th If any user's MSE c,k ≤ε th If the channel estimation is accurate, the user is assigned to the NOMA user group.
[0024] If any user's MSE c,k >ε th If so, it will be assigned to the traditional power allocation group;
[0025] MSE threshold ε th The tolerance of the system to channel estimation error is determined through simulation parameter tuning.
[0026] As a preferred embodiment of the hybrid optimization method for UAV multi-input multi-output communication networks described in this invention, the optimization of LED transmitter selection includes setting each UAV to have 4 LED transmitters and each user to have 4 PD receivers;
[0027] The optimal PD combination for each user is selected by branch and bound method. The branching method decomposes the problem into multiple sub-problems, and each sub-problem corresponds to the selection of an LED combination.
[0028] The delimiting method involves enumerating possible PD combinations for each user, setting the currently selected PD combination as optimal, estimating the maximum signal-to-interference-plus-noise ratio (SNR) to be achieved, and pruning inefficient branches based on this. If the upper bound of any combination is lower than the current optimal solution, it is skipped directly; otherwise, the current optimal solution is updated. The optimal LED combination is selected through the branch-and-bound method to reduce the number of redundantly activated LEDs / PDs.
[0029] As a preferred embodiment of the hybrid optimization method for UAV multiple-input multiple-output communication networks described in this invention, the power allocation includes allocating different powers according to the user's channel gain in the NOMA group;
[0030] The channel gains of users within the group are sorted, and the sorting result is h1≤h2≤…≤h k , where h k The channel gain for the k-th user;
[0031] The system uses an allocation method for power distribution. For a two-user scenario, if the total power of the system is P, the power allocated to the user with the earlier channel ranking result is αP, and the power allocated to the user with the later channel ranking result is (1-α)P, where α∈(0.5,1). Energy efficiency is improved by dynamically allocating power in the NOMA group.
[0032] As a preferred embodiment of the hybrid optimization method for UAV multiple-input multiple-output communication networks described in this invention, the step of maximizing the total system rate includes using the ADMM algorithm to jointly optimize LED selection, power allocation, and precoding matrix to maximize the total system rate.
[0033] Joint optimization is performed using the ADMM algorithm, with the objective function being:
[0034] Where, ξ k and ξ i These are the SINR values for users in the power allocation group and the NOMA group, respectively. p and u N These represent the user sets of the traditional power allocation group and the NOMA group, respectively.
[0035] Increase the transmission rate per unit spectrum by maximizing the objective function.
[0036] Another objective of this invention is to provide a hybrid optimization system for multi-input multi-output communication networks of unmanned aerial vehicles (UAVs).
[0037] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a hybrid optimization system for UAV multi-input multi-output communication networks, comprising: an initialization module, a user grouping module, an LED selection optimization module, a power allocation module, and a joint optimization module;
[0038] The initialization module is based on a hybrid framework of cluster reconstruction, in which a multi-cell, multi-UAV VLC system is deployed for scene setting and parameter configuration.
[0039] The user grouping module sets each transmitter to follow the Lambertian radiation distribution, calculates the user channel gain based on channel state information, and groups users accordingly.
[0040] The LED selection optimization module uses a branch and bound method to optimize the selection of LED transmitters;
[0041] The power allocation is performed based on the user channel gain according to the result of the user grouping.
[0042] The joint optimization module uses the alternating direction multiplier method to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system speed.
[0043] The present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the hybrid optimization method for UAV multiple input multiple output communication networks.
[0044] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the hybrid optimization method for UAV multiple-input multiple-output communication networks.
[0045] The beneficial effects of this invention are as follows: This invention introduces a dynamic user packet strategy based on channel conditions in UAV communication networks. By combining non-orthogonal multiple access (NOMA) and traditional power allocation technology, it can flexibly adjust the resource allocation scheme according to the differences in user channel states, significantly improving spectrum efficiency and system throughput, and realizing dynamic user packet and resource allocation optimization.
[0046] This invention employs a branch-and-bound method for LED selection, dynamically selecting necessary antennas for communication, thereby reducing system complexity and energy consumption.
[0047] This invention proposes a joint optimization framework based on the Alternating Direction Multiplier Method (ADMM), which can simultaneously solve non-convex optimization problems such as LED selection, power allocation, and precoding. By improving real-time performance through this joint optimization framework, a globally optimal solution is achieved, significantly enhancing the system's real-time performance and resource utilization efficiency.
[0048] This invention introduces a user grouping strategy based on mean square error (MSE), which can dynamically adjust the grouping scheme according to the accuracy of user channel estimation. This can better adapt to the rapid changes in channel state in UAV communication, reduce the impact of channel estimation error on system performance, and improve the reliability and stability of communication.
[0049] This invention effectively solves the interference problem in multi-cell UAV communication by combining dynamic user grouping and NOMA technology, enabling more efficient interference management in multi-cell environments and improving the communication quality of edge users.
[0050] This invention significantly reduces the computational complexity of the optimization problem by combining the branch and bound method with ADMM, thereby reducing the consumption of computing resources while ensuring performance, making the solution applicable to large-scale UAV networks. Attached Figure Description
[0051] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0052] Figure 1 is a flowchart of the overall process of a hybrid optimization method for UAV multiple-input multiple-output communication networks provided in an embodiment of the present invention.
[0053] Figure 2 is a diagram of a multi-cell, multi-UAV VLC system architecture of a hybrid optimization method for UAV multiple-input multiple-output communication networks provided in an embodiment of the present invention.
[0054] Figure 3 is a schematic diagram of user grouping based on MSE in a hybrid optimization method for UAV multiple-input multiple-output communication networks provided in an embodiment of the present invention. Detailed Implementation
[0055] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.
[0056] Example 1, referring to Figures 1-3, is the first embodiment of the present invention. This embodiment provides a hybrid optimization method for UAV multiple-input multiple-output communication networks, including:
[0057] S1: A hybrid framework based on cluster reconstruction, in which a multi-cell, multi-UAV VLC system is deployed for scene setting and parameter configuration.
[0058] Specifically, the architecture diagram of the multi-cell, multi-drone VLC system is shown in Figure 2.
[0059] The scenario setup includes deploying a multi-cell, multi-drone system, with each drone equipped with multiple LED transmitters and users randomly distributed and equipped with multiple PD receivers.
[0060] The parameter configuration includes initialization of Channel State Information (CSI) acquisition parameters, MSE threshold (ε), power allocation ratio (α), LED selection range, etc.
[0061] S2: Set each transmitter to follow the Lambertian radiation distribution, calculate the user channel gain based on channel state information, and group users.
[0062] It should be noted that each LED is assumed to follow the Lambertian radiation distribution, and the user channel gain h is calculated based on channel state information. c,i,k,r :
[0063] Where m is the Lambertian exponent, m = -ln(2) / ln(cos(Φ) 1 / 2 )), Φ 1 / 2 It is the half-power angle, A r It is the effective area of PD, d c,i,k,r φ is the distance between the i-th LED in the c-th group and the i-th PD of the k-th user. c,i,k,r and ψ c,i,k,r These are the incident angle and radiation angle of the i-th PD of the k-th user, respectively, T s (ψ c,i,k,r ) is the gain of the optical filter, g(ψ) c,i,k,r ) is the gain of the optical concentrator.
[0064] It should be noted that, based on the user's Channel State Information (CSI), which includes a set of characteristic parameters such as channel gain and path loss between the user and the transmitter, the user's current channel is estimated, and users with similar estimation error levels are grouped together.
[0065] User grouping is based on MSE, and the formula for calculating MSE is:
[0066] in, It is the estimated channel, MSE c,i,k,r To approximate the estimation error level, F is the Frobenius norm, and N... rTo receive the number of PDs, N c,t This refers to the number of subcarriers.
[0067] Set the predefined MSE threshold ε th If any user's MSE c,k ≤ε th If the channel estimation is accurate, the user is assigned to the NOMA user group.
[0068] If any user's MSE c,k >ε th If so, it will be assigned to the traditional power allocation group;
[0069] MSE threshold ε th Based on the system's tolerance for channel estimation errors, the value range ε is determined through simulation parameter tuning. th ∈[10 -4 10 -2 The specific value depends on the noise power and channel quality.
[0070] In an alternative embodiment, user groups are not limited to MSE-based groups, but can also be grouped based on other channel characteristics or user requirements.
[0071] S3: The branch-and-bound method is used to optimize the selection of LED transmitters.
[0072] In UAV communication systems, LED selection directly impacts signal quality, interference management, and system power consumption. Branch-and-bound is a commonly used combinatorial optimization method suitable for selecting the optimal solution from multiple transmit-receive combinations, and has been widely applied to resource scheduling and LED selection problems in wireless communication. This patent employs the branch-and-bound method for LED selection to find the optimal combination of transmit LEDs and receive PDs within acceptable complexity.
[0073] Furthermore, each drone is configured to have 4 LED transmitters, and each user has 4 PD receivers;
[0074] By using the branch-and-bound method, the optimal PD combination for each user is selected to reduce interference and improve signal quality.
[0075] The branching method involves breaking down the problem into multiple sub-problems, with each sub-problem corresponding to the selection of a combination of LEDs.
[0076] The delimiting method involves enumerating possible PD combinations for each user, setting the currently selected PD combination as optimal, estimating the maximum signal-to-interference-plus-noise ratio (SNR) to be achieved, and pruning inefficient branches based on this. If the upper bound of any combination is lower than the current optimal solution, it is skipped directly; otherwise, the current optimal solution is updated. The optimal LED combination is selected through the branch-and-bound method to reduce the number of redundantly activated LEDs / PDs.
[0077] It should be noted that the inefficient branches refer to the PD combinations in the branch and bound method whose performance cannot be better than the currently known optimal solution after estimating the upper bound value; these branches will be directly pruned (skipped) to reduce computational complexity.
[0078] The maximum signal-to-interference-plus-noise ratio (SINR) is not estimated as a predetermined fixed value, but rather calculated dynamically. For each candidate PD combination, its upper bound is estimated by calculating the maximum SINR when interference is completely eliminated.
[0079] The upper bound refers to the theoretical upper limit of the best performance that a certain PD combination can achieve (the maximum SINR when interference is completely eliminated).
[0080] The current optimal solution is the best feasible solution found by the algorithm during the search process (i.e., the solution with the highest actual SINR among the verified PD combinations). This solution will be continuously updated as the search progresses (when a better branch is found).
[0081] In an alternative embodiment, LED selection is not limited to the branch and bound method; other optimization algorithms such as greedy algorithms or genetic algorithms may also be used.
[0082] S4: Based on the results of the user grouping, power is allocated through the user channel gain.
[0083] It should be noted that within the NOMA group, different power is allocated based on the user's channel gain; the channel gains of users within the group are sorted, i.e., the sorting result is h1≤h2≤…≤h k , where h k The channel gain for the k-th user;
[0084] The system employs either a fixed-ratio power allocation method or an optimized allocation method. Taking the fixed-ratio method as an example, for a two-user scenario, if the total system power is P, then the power allocated to the user ranked first in the channel ranking is αP, and the power allocated to the user ranked second in the channel ranking is (1-α)P, where α∈(0.5,1). In the NOMA group, the signal-to-interference-plus-noise ratio (SINR) of a user can be expressed as:
[0085] Where γ is the responsivity of the PD, μ is the conversion factor of the LED, and p i T is the power allocated to user i. i H is the LED selection matrix for user i. i Let A be the channel matrix for user i. c F is the LED selection matrix for cell c. i It is the precoding matrix of user i. It is the noise power at the receiving end, u pIt is the set of users in the p-th cell that is different from the current cell, and e is the natural base.
[0086] In an alternative embodiment, power allocation is not limited to NOMA, but can also be combined with other multiple access technologies such as OFDMA or TDMA; energy efficiency is improved by dynamically allocating power in NOMA groups.
[0087] S5: Use the alternating direction multiplier method to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system speed.
[0088] The ADMM algorithm is used to jointly optimize LED selection, power allocation, and the precoding matrix to maximize the overall system speed. ADMM decomposes complex non-convex optimization problems into multiple subproblems, which are solved iteratively in parallel, effectively improving convergence speed, and is particularly suitable for high-density user scenarios. By maximizing the weighted logarithmic sum of the SINR of all users, the overall system throughput is effectively improved.
[0089] It should be noted that the ADMM algorithm is used to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system speed;
[0090] Joint optimization is performed using the ADMM algorithm, with the objective function being:
[0091] Where, ξ k and ξ i These are the SINR values for users in the power allocation group and the NOMA group, respectively. p and u N These represent the user sets of the traditional power allocation group and the NOMA group, respectively.
[0092] Because resources (such as power, beam, and LEDs) are allocated on demand among users, the system can maximize resource utilization efficiency while meeting the communication quality requirements of different users and avoiding resource waste.
[0093] Specifically, the transmission rate per unit spectrum is increased by maximizing the objective function; energy efficiency is improved by dynamically allocating power in the NOMA group (less power to users with strong channels and more power to users with weak channels); and the number of redundantly activated LEDs / PDs is reduced by selecting the optimal LED combination through the branch-and-bound method.
[0094] Example 2, the second embodiment of the present invention, provides a hybrid optimization method for UAV multi-input multi-output communication networks. To verify the beneficial effects of the present invention, scientific demonstration is carried out through experiments.
[0095] Suppose there are two cells (cell 1 and cell 2) in an indoor environment, each with 4 drones and 8 users (equipped with 4 device detectors). The drones are distributed in different locations, and the users are randomly distributed indoors.
[0096] (1) User Grouping: Users are divided into NOMA group and traditional power allocation group by calculating their MSE. For example, users 1, 3, 6 and 7 in cell 1 have smaller MSEs and are assigned to the NOMA group; users 2, 4, 5 and 8 have larger MSEs and are assigned to the traditional power allocation group. The user grouping is shown in Figure 3.
[0097] (2) LED selection: The branch-and-bound method is used to select the optimal PD combination for each user. For example, user 1 selects PD1 and PD2, and user 2 selects PD3 and PD4 to reduce interference and improve signal quality.
[0098] (3) Power allocation: In the NOMA group, user 1 has a stronger channel gain and is allocated lower power; user 2 has a weaker channel gain and is allocated higher power.
[0099] (4) Joint optimization: The ADMM algorithm is used to jointly optimize LED selection, power allocation and precoding matrix to ensure the resource utilization efficiency of the system in high-density user scenarios.
[0100] Example 3 is the third embodiment of the present invention, which differs from the previous two embodiments in that:
[0101] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.
[0102] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-included system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.
[0103] More specific examples of computer-readable media (a non-exhaustive list) include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.
[0104] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.
[0105] Example 4 is the fourth embodiment of the present invention. This embodiment provides a hybrid optimization system for UAV multiple-input multiple-output communication networks, including: an initialization module, a user grouping module, an LED selection optimization module, a power allocation module, and a joint optimization module.
[0106] The initialization module is based on a hybrid framework of cluster reconstruction, in which a multi-cell, multi-UAV VLC system is deployed for scene setting and parameter configuration;
[0107] The user grouping module sets each transmitter to follow the Lambertian radiation distribution, calculates the user channel gain based on channel state information, and groups users accordingly.
[0108] The LED selection optimization module uses a branch and bound method to optimize the selection of LED transmitters.
[0109] Power allocation is performed based on the user grouping results and through the user channel gain.
[0110] The joint optimization module uses the alternating direction multiplier method to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system speed.
[0111] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A hybrid optimization method for unmanned aerial vehicle multiple-input multiple-output communication networks, characterized by: include, Based on a cluster reconstruction hybrid framework, a multi-cell, multi-UAV VLC system is deployed in the framework for scene setting and parameter configuration; Each transmitter is configured to follow the Lambertian radiation distribution. User channel gain is calculated based on channel state information, and users are grouped. The branch-and-bound method is used to optimize the selection of LED transmitters; Based on the results of the user grouping, power is allocated using the user channel gain; The alternating direction multiplier method is used to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system rate. 2.The UAV MIMO communication network hybrid optimization method of claim 1, wherein: The scenario setup includes deploying a multi-cell, multi-drone system, with each drone equipped with multiple LED transmitters and users randomly distributed and equipped with multiple PD receivers; The parameter configuration includes the acquisition parameters for initializing channel state information, MSE threshold, power allocation ratio, and LED selection range. 3.The UAV MIMO communication network hybrid optimization method of claim 2, wherein: The calculation of user channel gain includes, Set each LED to follow Lambertian radiation distribution, calculate user channel gain h based on channel state information c,i,k,r : where m is the Lambertian exponent, m = -ln(2) / ln(cos(Φ 1 / 2 )), Φ 1 / 2 is the half-power angle, A r is the effective area of the PD, d c,i,k,r is the distance between the ith LED of the cth group and the ith PD of the kth user, φ c,i,k,r and ψ c,i,k,r are the incidence and emission angles of the ith PD of the kth user, respectively, T s (ψ c,i,k,r ) is the gain of the optical filter, and g(ψ c,i,k,r ) is the gain of the optical concentrator. 4.The UAV MIMO communication network hybrid optimization method of claim 3, wherein: The user grouping includes estimating the user's current channel based on the user's channel state information, which includes a set of characteristic parameters such as channel gain and path loss between the user and the transmitter, and grouping users with similar estimation error levels into the same group. The user group is based on MSE, and the calculation formula of MSE is: wherein, is the estimated channel, MSE c,i,k,r is the close estimate error level, F is the Frobenius norm, N r is the number of received PDs, N c,t is the number of subcarriers; Setting a predefined MSE threshold ε th If the MSE of any user c,k ≤ε th , the channel estimation is considered accurate and the user is assigned to the NOMA user group. If MSE of any user c,k > ε th then divide to legacy power allocation group; MSE threshold ε th According to the system tolerance to channel estimation error, the simulation is adjusted to determine. 5.The UAV MIMO communication network hybrid optimization method of claim 4, wherein: The optimized selection of LED transmitters includes setting each drone to have 4 LED transmitters and each user to have 4 PD receivers; The optimal PD combination for each user is selected by branch and bound method. The branching method decomposes the problem into multiple sub-problems, and each sub-problem corresponds to the selection of an LED combination. The delimiting method involves enumerating possible PD combinations for each user, setting the currently selected PD combination as optimal, estimating the maximum signal-to-interference-plus-noise ratio (SNR) to be achieved, and pruning inefficient branches based on this. If the upper bound of any combination is lower than the current optimal solution, it is skipped directly; otherwise, the current optimal solution is updated. The optimal LED combination is selected through the branch-and-bound method to reduce the number of redundantly activated LEDs / PDs. 6.The UAV MIMO communication network hybrid optimization method of claim 4, wherein: The power allocation includes, within the NOMA group, allocating different powers based on the user's channel gain; The channel gains of the users in the group are sorted in ascending order as h1≤ h2≤ …≤ hk≤ …≤ hn k where h k is the channel gain of the kth user. The system uses an allocation method for power distribution. For a two-user scenario, if the total power of the system is P, the power allocated to the user with the earlier channel ranking result is αP, and the power allocated to the user with the later channel ranking result is (1-α)P, where α∈(0.5,1). Energy efficiency is improved by dynamically allocating power in the NOMA group.
7. The unmanned aerial vehicle multiple-input multiple-output communication network hybrid optimization method of claim 4, wherein: The step of maximizing the overall system rate includes using the ADMM algorithm to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system rate. The joint optimization is performed using the ADMM algorithm, with the objective function being: wherein, ξ k and ξ i are the SINRs of the power allocation group and the NOMA group users, respectively, u p and u N represent the user sets of the conventional power allocation group and the NOMA group, respectively. Increase the transmission rate per unit spectrum by maximizing the objective function.
8. A hybrid optimization system for a UAV MIMO communication network, applying the hybrid optimization method for a UAV MIMO communication network according to any one of claims 1-7, characterized in that, include: Initialization module, user grouping module, LED selection optimization module, power allocation and joint optimization module; The initialization module is based on a hybrid framework of cluster reconstruction, in which a multi-cell, multi-UAV VLC system is deployed for scene setting and parameter configuration. The user grouping module sets each transmitter to follow the Lambertian radiation distribution, calculates the user channel gain based on channel state information, and groups users accordingly. The LED selection optimization module uses a branch and bound method to optimize the selection of LED transmitters; The power allocation is performed based on the user channel gain according to the result of the user grouping. The joint optimization module uses the alternating direction multiplier method to jointly optimize LED selection, power allocation, and precoding matrix to maximize the overall system speed. 9.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-8 when the computer program is executed by the processor. When the processor executes the computer program, it implements the steps of the hybrid optimization method for UAV multi-input multi-output communication networks as described in any one of claims 1 to 7.
10. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the hybrid optimization method for UAV multi-input multi-output communication networks as described in any one of claims 1 to 7.
