Sum rate maximization method for UAV-assisted NOMA system under hardware damage conditions

Abstract

The present invention claims protection for a method for maximizing the sum rate of a non-orthogonal multiple access (NOMA) system assisted by an unmanned aerial vehicle (UAV) under hardware damage conditions, which belongs to the field of wireless resource management. Taking into account the hardware damage in actual scenarios, the system sum rate is maximized by jointly optimizing the position deployment of the UAV and the power allocation of the user. The method first decomposes the system sum rate maximization problem into the UAV position optimization problem and the user power allocation sub-problem, and solves the sub-problems respectively through the successive convex approximation method (SCA) and the Lagrange dual transformation and the quadratic transformation method, and finally solves the system sum rate by iteration. The result of the present invention improves the sum rate of the UAV-assisted NOMA system, ensures the user's service quality, and has strong practicality.

Description

Technical Field

[0001] The present invention belongs to the technical field of wireless resource management, and specifically relates to a method for allocating power resources of an unmanned aerial vehicle (UAV)-assisted base station to a downlink multi-user non-orthogonal frequency division multiple access (NOMA) system under conditions of hardware damage. Background Art

[0002] Compared to the currently commercially available 5th generation mobile communications, the future 6th generation mobile communication technology will require higher performance metrics, such as access to ultra-large-scale IoT devices and ultra-fast information transmission rates. Non-orthogonal frequency division multiple access (NOMA), a promising candidate for next-generation communication networks, has attracted extensive research in academia due to its ability to effectively improve system capacity and spectrum efficiency. Compared to traditional orthogonal multiple access (OMA), NOMA's primary advantage lies in its ability to multiplex power resources at the transmitter, allowing multiple users to simultaneously share the same time, frequency, and code domain resources. At the receiver, successive interference cancellation (SIC) is used to decode the received signal from the superimposed signals.

[0003] At the same time, unmanned aerial vehicle (UAV) communications, a recently emerging technology, can help NOMA achieve even better performance due to their high maneuverability and low overhead. Drones are often deployed as flying base stations or mobile relays to help the system achieve higher throughput and wider coverage. Compared to traditional terrestrial relays, drone relays offer greater flexibility and can be moved to avoid obstacles or severe shadowing that can degrade system performance.

[0004] In recent years, in order to combine the success of NOMA technology and drone communication technology, there have been many studies on the power resource allocation of NOMA downlink multi-user networks, but most of them assume in advance that the system has ideal hardware conditions, and the decoding sequence of SIC is also pre-supposed to be sorted. However, this is obviously not in line with the actual scenario. In actual application scenarios, the position of the drone can be changed. As the position of the drone changes, the quality of the channel conditions will also change accordingly, and the hardware in the system is often affected by hardware damage such as in-phase orthogonal imbalance, nonlinear amplification noise, and RF circuit noise. Therefore, the present invention takes the user's quality of service (QoS) and the drone's flight area as restrictions, and proposes a power resource allocation method for maximizing the downlink user sum rate in the NOMA system of drone-assisted communication under hardware damage conditions. Summary of the Invention

[0005] The present invention aims to solve the above problems in the prior art. It proposes a method for maximizing the sum rate of a drone-assisted NOMA system under hardware damage conditions. The technical solution of the present invention is as follows:

[0006] A method for maximizing the sum rate of a UAV-assisted NOMA system under hardware damage conditions comprises the following steps:

[0007] Step 1) Initialize the noise power, number of users, user power allocation, base station transmit power, drone transmit power, base station to drone link hardware damage level, drone to user n link hardware damage level, base station location, user location, drone initial location, decoding order sequence, number of iterations, tolerance error threshold, calculate the initial sum rate, and establish the system sum rate maximization problem.

[0008] Step 2) Decompose the system and rate maximization problem into two sub-problems: UAV position optimization problem and power allocation problem;

[0009] Step 3) Solve the UAV position optimization problem by variable substitution: convert the objective function of the UAV position optimization problem, a mixed non-convex integer optimization problem, into a convex function, use SCA technology to transform the non-convex constraints into convex constraints, and convert the UAV position optimization problem into a standard convex optimization problem; use the interior point method to solve the UAV position optimization problem.

[0010] Step 4) Solve the power allocation problem: Use the Lagrange dual transformation method and the quadratic transformation method to transform the objective function of the power allocation optimization problem into a convex function; adopt the Lagrange transformation method to introduce the approximate variable Υ=(Υ1,...,Υ n), then according to the quadratic transformation rule, the approximate variable can be written Finally, by updating the parameter Υ, And use the interior point method to solve the power allocation optimization problem;

[0011] Step 5) Substitute the UAV position and power obtained from step 3) UAV position optimization and step 4) power allocation optimization into the downlink sum rate formula to update the system downlink sum rate;

[0012] Step 6) Repeat step 3) UAV position optimization, step 4) power allocation optimization, and step 5) system downlink and rate update, iteratively update the UAV position and power until the convergence condition is met, and give the system and rate.

[0013] Furthermore, in step 1), the noise power σ between the drone and the base station is initialized R 2 , the noise power between the UAV and user n The number of users is N, and the user power allocation P = (p1, ..., p n ), P s Indicates the base station transmission power, P s max Indicates the maximum transmission power of the base station, P r Indicates the UAV transmission power, P r max Denotes the maximum transmission power of the UAV, and defines η sr Indicates the total hardware damage from the base station to the drone receiving end, subject to where κ sr Indicates the overall hardware damage level between the base station and the drone, meeting where κ s ,κ r Respectively represent the hardware damage suffered at the base station and the drone, η rn Indicates the total hardware damage from the drone to the user end, subject to where κ rn Indicates the overall hardware damage level between the drone and the user end, satisfying where κ r ,κ n They represent the hardware damage suffered by the drone and user n respectively; the tolerance is set to δ, and the 3D coordinates of the base station and the user are (x s ,y s ,0),(x n ,y n ,0),n=1,...,N; the initial position of the UAV is set to (x,y,H), where H represents the altitude of the UAV, and the channel coefficient from the base station to the UAV is defined as h sr, the channel coefficient from the UAV to the user terminal n is h rn ;Decoding order sequence satisfy where α k,n is a binary variable, the number of iterations l, the calculation system downlink rate R sum , establish the system downlink and rate maximization problem P1:

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

[0020]

[0021]

[0022]

[0023]

[0024]

[0025] Constraint C1a indicates that the user power is non-negative; Constraint C1b indicates that the sum of the user powers is less than or equal to the UAV base station transmit power; Constraint C1c indicates the UAV transmit power limit; Constraint C1d indicates the base station transmit power limit; Constraint C1e indicates that the minimum rate of user n is higher than the quality of service threshold Constraint C1f represents user fairness; Constraint C1g represents that user n cannot regard its own signal as interference; Constraint C1h represents the decoding order relationship between any two users k and n; Constraint C1i represents that for any two users n and k, one user must be able to decode and eliminate the other user; Constraint C1j represents that when a k,n = 1, the distance between user n and the drone is greater than the distance between user k and the drone.

[0026] Furthermore, in step 3), the initialization power is introduced into problem P1 to obtain the UAV position optimization problem P2. Now let H 2 +(xx k ) 2 +(yy k ) 2 =d rk,H 2 +(xx n ) 2 +(yy n ) 2 =d rn , the specific form is as follows:

[0027]

[0028]

[0029]

[0030]

[0031]

[0032]

[0033] Constraint C2a means that user n cannot regard its own signal as interference; Constraint C2b means the decoding order relationship between any two users k and n; Constraint C2c means that for any two users n and k, one user must be able to decode and eliminate the other user; Constraint C2d means that when a k,n = 1, the distance between user n and the drone is farther than the distance between user k and the drone. Constraint C2e indicates that the rate of any user n should be greater than its preset threshold.

[0034] Furthermore, the step 3) introduces the variable R=(R1,...,R n ),T=(t1,...,t n ) to replace the objective function and use the SCA technique to scale the constraint C2b. P2 can be transformed into a standard convex optimization problem P3, which is as follows:

[0035]

[0036]

[0037]

[0038]

[0039]

[0040]

[0041]

[0042]

[0043]

[0044] Constraint C3a means that user n cannot treat its own signal as interference; constraint C3b means that for any two users n and k, one user must be able to decode and eliminate the other user; constraint C3c means that the binary decision variable α k,n The value range constraint of C3d represents the value range of binary decision variable α k,n Scaling; Constraint C3e means that when a k,n = 1, the distance between user n and the drone is farther than the distance between user k and the drone; the constraint C3f means that for the variable T = (t1, ..., t n ) scaling constraint; Constraint C3g represents the scaling constraint for the variable R = (R1, ..., R2); Constraint C3h represents that the rate of any user n should be greater than its preset threshold At this point, the interior point method can be used to solve the position of the drone.

[0045] Furthermore, in step 4), the position of the UAV in step 3) is substituted into problem P1 to obtain the user power allocation problem P4, which is specifically formulated as follows:

[0046]

[0047]

[0048]

[0049]

[0050]

[0051] Constraint C4a indicates that the sum of user powers should not exceed the transmit power; constraint C4b indicates user fairness; constraint C4c indicates that user power is non-negative; constraint C4d indicates that the n-terminal rate of any user should be greater than its preset threshold By performing Lagrange transformation on the objective function, the approximate variable Υ=(Υ1,...,Υ n ) replace the complex fraction part and introduce approximate variables according to the quadratic transformation regulations P4 can be transformed into a standard convex optimization problem P5, which has the following specific form:

[0052]

[0053]

[0054]

[0055]

[0056]

[0057] in By order Can be obtained:

[0058]

[0059] Similarly set up Can be obtained:

[0060]

[0061] P=(p1,...,p n ) represents the user power allocation,

[0062]

[0063] Finally, we have:

[0064]

[0065] Constraint C5a indicates that the sum of user powers should not exceed the transmit power; constraint C5b indicates user fairness; constraint C5c indicates that user power is non-negative; constraint C5d indicates that the n-terminal rate of any user should be greater than its preset threshold At this point, the interior point method can be used to solve the user power allocation optimization problem.

[0066] Furthermore, in step 5), the drone position obtained in step 3 and the public power distribution obtained in step 4 are substituted into the following formula:

[0067] To calculate the downlink and rate for update.

[0068] Furthermore, in step 6), repeat step 3) drone position optimization, step 4) user power allocation optimization, and step 5) system downlink and rate update to iteratively update the drone position and power. l Indicates that when the lth iteration, the value of the sum rate R is taken until R is satisfied l+1 -R l When <δ, it means that the difference between the l+1th sum rate and the lth sum rate is within the tolerance δ, and the convergence condition is reached; based on the proposed user power allocation method and UAV position optimization method, the final output is the UAV deployment position (x, y) under the maximum downlink sum rate, the SIC decoding sequence A, the user allocation power P, and the system sum rate R sum .

[0069] The advantages and beneficial effects of the present invention are as follows:

[0070] The present invention considers the power resource allocation problem based on user and rate maximization in the downlink transmission NOMA system.

[0071] 1. Most previous studies on the downlink and rate of drone-assisted NOMA systems have considered ideal hardware conditions. In practice, communication equipment is often affected by hardware damage. Unlike most previous studies, this paper introduces different levels of hardware damage to each node in the communication. It is worth noting that, for the sake of being closer to reality, the noise power received at the drone node and the user-end node is also different.

[0072] 2. Previous studies often only focus on theoretical implementation, while ignoring the problem that high complexity cannot be easily implemented. The complexity of the method proposed in this invention is K is the number of iterations, N is the problem size, and ε is the tolerance error of the interior point method. Therefore, the method proposed in this invention is feasible. This method decomposes the original problem into two sub-problems: user power allocation and drone position optimization. Since the optimization problem is not a standard convex optimization problem, this invention introduces approximate auxiliary variables and decoding order variables. The sub-problems are converted to standard convex problems through the successive convex approximation (SCA) method, the Lagrange dual transformation method, and the quadratic transformation method. The sub-problems are solved using the interior point method, and the original problem is solved through an iterative method.

[0073] 3. Regarding the SIC decoding order sorting problem, most of the previous studies on drone-assisted NOMA systems and rates have been to preset the decoding order in advance. However, the mobility of drones poses a challenge to this method, because as the position of the drone changes, the channel conditions between it and the user also change accordingly. At this time, assuming the decoding sequence in advance is too ideal and does not conform to the actual communication situation. The present invention introduces a SIC decoding order sequence variable, which no longer assumes the channel from the drone to multiple users and sorts it in advance. In addition, it maximizes the system and rate while ensuring the minimum service quality of each user. The final results prove the effectiveness of this method. Compared with the NOMA scheme of deploying drones at the user geometric center and the orthogonal frequency division multiple access (OMA) scheme of deploying drones by traversing and searching for the optimal position, the system and rate are always the highest. The present invention also analyzes the impact of the number of access users on the system and rate, and has better practicality and feasibility. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 The present invention provides a preferred embodiment that provides a model diagram of a drone-assisted NOMA communication system under conditions of hardware damage.

[0075] Figure 2 It is a sum rate curve diagram of the present invention compared with the OMA solution under different hardware damage conditions.

[0076] Figure 3 This is the impact of the number of users on the system and rate in the method of the present invention.

[0077] Figure 4 It is a curve chart of the UAV deployment height compared with other solutions.

[0078] Figure 5 It is a schematic flow diagram of the present invention. DETAILED DESCRIPTION

[0079] The following will describe the technical solutions in the embodiments of the present invention in detail with reference to the accompanying drawings. The described embodiments are only a part of the embodiments of the present invention.

[0080] The technical solution of the present invention to solve the above technical problems is:

[0081] This implementation case is a power resource allocation method for maximizing downlink and rate in a UAV-assisted NOMA system under hardware damage conditions.

[0082] The specific implementation cases are as follows:

[0083] Step 1: Initialize the number of users to N. The base station and users are placed on the ground, and their height is assumed to be 0m. Their 3D coordinates are (x s ,y s ,0),(x n ,y n ,0),n=1,...,N. The initial position of the drone is set to (x,y,H), where H represents the altitude of the drone. For the convenience of exploration, this paper assumes that the altitude of the drone is fixed. Set the tolerance δ. The superposition signal sent by the base station is where p n represents the power allocated to user n, θ n Represents the symbol information sent to user n, satisfying E(|θ n | 2 ) = 1. It is assumed that there is no direct link between the base station and the user due to the severe shadowing effect. The link from the drone to the base station and the ground user is a line-of-sight link. The entire communication process is divided into two time slots. In the first time slot: the base station sends superimposed information to the drone. The channel modeling from the base station to the drone and from the drone to the ground user n is:

[0084]

[0085] Where β0 represents the unit distance channel power gain.sr ,d rn They represent the distances between the UAV and the base station and user n respectively.

[0086] The signal received on the drone side is: Define η sr Indicates the total hardware damage from the base station to the drone receiving end, subject to Among them, P s represents the power transmitted by the base station, κ sr Indicates the overall hardware damage level between the base station and the drone, meeting κ s ,κ r Represents the hardware damage suffered at the base station and the drone, n r represents additive Gaussian white noise, obeying The UAV adopts the amplify-and-forward protocol, and its amplification gain G is: P r Indicates the power transmitted by the drone. In the second time slot: the drone amplifies the signal and forwards it to user n. The final signal received by user n is:

[0087] η rn represents the total hardware damage from the drone to the user end, subject to where κ rn represents the overall hardware damage level between the UAV and the user terminal n, P r Indicates the UAV's transmission power, satisfying κ r ,κ n Respectively represent the hardware damage suffered by the drone and user n, represents additive Gaussian white noise, obeying Initialize the number of iterations l.

[0088] SIC technology decodes different users based on different channel gains. Since the position of the drone is not fixed, the variable To express the SIC decoding order. When α k,n =1, indicating that user k has a stronger channel power gain, and the signal of user k is regarded as noise when decoding the signal of user n; in other cases, α k,n = 0. Decoding order sequence Specifically defined as: Because α k,n is an integer binary variable and can be rewritten as:

[0089] where d k ,d n Respectively represent the distance between the drone and user k, n, satisfying The signal-to-noise ratio at user n is:

[0090] The system downlink and rate maximization problem P1 can be formulated as:

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102] Constraint C1a indicates that the user power is non-negative; Constraint C1b indicates that the sum of the user powers is less than or equal to the UAV base station transmit power; Constraint C1c indicates the UAV transmit power limit; Constraint C1d indicates the base station transmit power limit; Constraint C1e indicates that the minimum rate of user n is higher than the quality of service threshold Constraint C1f represents user fairness; Constraint C1g represents that user n cannot regard its own signal as interference; Constraint C1h represents the decoding order relationship between any two users k and n; Constraint C1i represents that for any two users n and k, one user must be able to decode and eliminate the other user; Constraint C1j represents that when a k,n = 1, the distance between user n and the drone is greater than the distance between user k and the drone.

[0103] It is easy to verify that the objective function increases monotonically with respect to the power of the UAV and the base station. Therefore, in the present invention, the transmission power P of the base station and the UAV is taken as r max , P s max The maximum value is used to solve the problem. Since this problem is a non-convex optimization problem, it is decomposed into the UAV position optimization problem and the power allocation problem. The optimization variables (x, y), A and P = (p1, ..., p n) for block-by-block solution.

[0104] Step 2: Decompose the system and rate maximization problem into two sub-problems: drone position optimization problem and power allocation problem.

[0105] Step 3: Solve the drone position optimization problem. The drone position optimization problem P2 is a fixed user power variable block P, which is substituted into the system downlink and rate maximization problem P1 to obtain:

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112] Constraint C2a means that user n cannot regard its own signal as interference; Constraint C2b means the decoding order relationship between any two users k and n; Constraint C2c means that for any two users n and k, one user must be able to decode and eliminate the other user; Constraint C2d means that when a k,n = 1, the distance between user n and the drone is farther than the distance between user k and the drone. Constraint C2e indicates that the rate of any user n should be greater than its preset threshold.

[0113] For the non-convex constraint C2b, its equivalent form is: right Through the first-order Taylor expansion, it can be further expanded to: Since the objective function is not strictly concave, a non-negative approximate variable R=(R1,...,R n ),T=(t1,...,t n ) replaces the objective function, the UAV position optimization problem can be rewritten as P3:

[0114]

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123] Constraint C3a means that user n cannot treat its own signal as interference; constraint C3b means that for any two users n and k, one user must be able to decode and eliminate the other user; constraint C3c means that the binary decision variable α k,n The value range constraint of C3d represents the value range of binary decision variable α k,n Scaling; Constraint C3e means that when a k,n = 1, the distance between user n and the drone is farther than the distance between user k and the drone; the constraint C3f means that for the variable T = (t1, ..., t n ) scaling constraint; Constraint C3g represents the scaling constraint for the variable R = (R1, ..., R2); Constraint C3h represents that the rate of any user n should be greater than its preset threshold

[0124] For the first term on the right side of the non-convex constraint C3f inequality, the SCA technique can be used to obtain:

[0125]

[0126] For the second term on the right side of the inequality

[0127]

[0128] in The difficulty in solving this problem is due to the multiplication of convex functions. By scaling, the non-convex term D rn D sr We can get:

[0129] in

[0130]

[0131]

[0132]

[0133] Therefore, the non-convex constraint C3f finally becomes a convex constraint and can be rewritten as:

[0134]

[0135] in

[0136] Similarly, for constraint C3h:

[0137] For constraint C3g, through simple mathematical operations and SCA technology, it can be rewritten as:

[0138] At this point, the UAV position optimization problem P2 is converted into a standard convex problem, which is solved by the interior point method to obtain the UAV position.

[0139] Step 4: Solve the power allocation problem. After substituting the drone position (x, y) and the decoding sequence A in step 3, the user power allocation optimization problem P4 can be established as:

[0140]

[0141]

[0142]

[0143]

[0144]

[0145] Constraint C4a indicates that the sum of user powers should not exceed the transmit power; constraint C4b indicates user fairness; constraint C4c indicates that user power is non-negative; constraint C4d indicates that the n-terminal rate of any user should be greater than its preset threshold

[0146] This subproblem is difficult to solve because its objective function is non-convex, so the objective function is Lagrange transformed and the approximate variable Υ=(Υ1,...,Υ n ) replace the complex fraction part and introduce approximate variables according to the quadratic transformation regulations Its objective function can be rewritten as: in

[0147] in By order The Υ update criterion can be obtained:

[0148]

[0149] For the fraction and part of f(P,Y), using the quadratic transformation, it can be rewritten as:

[0150]

[0151] Among them are:

[0152]

[0153] is an approximate variable, and The update criteria are:

[0154] Substitute it into The objective function can be obtained At this point, the sub-problem user power allocation is transformed into the standard convex problem P5, which is as follows:

[0155]

[0156]

[0157]

[0158]

[0159]

[0160] At this point, the user power allocation can be solved by using the interior point method.

[0161] Step 5: Substitute the UAV position obtained from the UAV position optimization problem in step 3 and the power allocation obtained from the power allocation optimization problem in step 4, and Calculate the downlink sum rate for update.

[0162] Step 6: Repeat the third step of drone position optimization, the fourth step of user power allocation optimization, and the fifth step of system downlink and rate update, iteratively updating the drone position and power, R l Indicates that when the lth iteration, the value of the sum rate R is taken until R is satisfied l+1 -R l When <δ, it means that the difference between the l+1th sum rate and the lth sum rate is within the tolerance δ, and the convergence condition is reached. Based on the proposed user power allocation method and drone position optimization method, the final output is the drone deployment position (x, y) under the maximum downlink sum rate, the SIC decoding sequence A, the user allocation power P, and the system sum rate R sum .

[0163] This invention takes into account the potential for hardware damage to communication hardware in real-world scenarios, making it more applicable to real-world scenarios. It also introduces a variable SIC decoding order sequence, eliminating the need for pre-ordering assumptions about the channels from drones to multiple users. Furthermore, it maximizes the system sum rate while ensuring minimum quality of service for each user. In the context of the provided method, the present invention achieves the highest system sum rate compared to the baseline comparison solution. The present invention also analyzes the impact of the number of connected users on the system sum rate, demonstrating strong practicality and feasibility.

[0164] This embodiment is a NOMA system for drone-assisted communication and a power allocation method for maximizing the rate under the condition of hardware damage. In this system network, the system noise The horizontal position of the fixed base station is (0,0), and users are randomly distributed in a 300×300m area 100m away from the base station. 2 Within the ground range, the user's minimum service quality is β0=10 -3 It is assumed that the UAV is subject to the same level of hardware damage as the ground user.

[0165] In an embodiment, Figure 1 This paper provides a system model for drone-assisted NOMA communication in the presence of hardware impairments. The network consists of a base station, a rotorcraft, and N downlink users. Due to severe shadowing or obstruction, there is no direct link between the base station and the users. The entire system communication process is divided into two time slots. All communication nodes are equipped with a single antenna. Figure 2 is the sum of the speed and the power of the UAV P UAV Change curve, fixed drone height H = 100m, base station transmission power P s = 2w, 3 randomly distributed users. The sum rate of the proposed NOMA solution for optimizing drone location deployment and the solution for traversing and searching for the optimal drone location in the OMA solution, both of which are subject to varying degrees of hardware damage, increases with increasing drone transmit power. The proposed NOMA solution consistently outperforms the OMA solution in terms of sum rate compared to the solution for traversing and searching for the optimal drone location in the OMA solution. Figure 3 The sum rate curves for different numbers of users, with a fixed drone height H = 100m and a base station transmission power P s =2w, UAV transmission power P UAV= 0.03W. When the user power allocation meets the user's minimum quality of service, the system and speed will increase as the number of users increases. When the user power allocation is insufficient to meet user needs, the system and speed will decrease as the number of users increases. In addition, as the hardware damage level increases at equal intervals, the system and speed will decrease more and more as the number of users increases. Figure 4 The graph showing the sum rate at different drone altitudes, with a fixed base station transmit power P s =2w, UAV transmission power P UAV = 0.03w, and the hardware damage level suffered is 0.01. As the drone deployment altitude increases, the distance between the drone and the user becomes farther, and the corresponding channel gain becomes worse. Therefore, the sum rate of the three schemes will show a downward trend. However, the sum rate of the NOMA scheme based on the proposed drone location optimization deployment is always better than the NOMA scheme based on the user geometric center deployment of the drone and the optimal drone deployment location scheme in the OMA scheme through traversal search.

[0166] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0167] The above embodiments should be understood as merely illustrating the present invention and not as limiting the scope of protection of the present invention. After reading the contents of the present invention, technicians may make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A method for maximizing the sum rate of a UAV-assisted NOMA system under hardware damage conditions, characterized in that: The following steps are involved: Step 1) Initialize the noise power, number of users, user power allocation, base station transmit power, drone transmit power, base station to drone link hardware damage level, drone to user n link hardware damage level, base station location, user location, drone initial location, decoding sequence, number of iterations, tolerance error threshold, calculate the initial sum rate, and establish the system sum rate maximization problem; Step 2) Decompose the system and rate maximization problem into two sub-problems: UAV position optimization problem and power allocation problem; Step 3) Solve the UAV position optimization problem by replacing variables: transform the objective function of the UAV position optimization problem, a mixed non-convex integer optimization problem, into a convex function, use SCA technology to transform the non-convex constraints into convex constraints, and transform the UAV position optimization problem into a standard convex optimization problem; use the interior point method to solve the UAV position optimization problem; Step 4) Solve the power allocation problem: Use the Lagrange dual transformation method and the quadratic transformation method to transform the objective function of the power allocation optimization problem into a convex function; use the Lagrange transformation method to introduce an approximate variable Υ=(Υ1,...,Υ n ) By replacing the variables, the approximate variables can be written according to the quadratic transformation rules. Finally, by updating the parameter Υ, And use the interior point method to solve the power allocation optimization problem; Step 5) Substitute the UAV position and power obtained from step 3) UAV position optimization and step 4) power allocation optimization into the downlink sum rate formula to update the system downlink sum rate; Step 6) Repeat step 3) UAV position optimization, step 4) power allocation optimization, and step 5) system downlink sum rate update, iteratively update the UAV position and power until the convergence condition is met, and give the system sum rate; In step 1), the noise power σ between the drone and the base station is initialized. R 2 , the noise power between the UAV and user n The number of users is N, and the user power allocation P = (p1, ..., p n ), P s Indicates the base station transmit power, Indicates the maximum transmission power of the base station, P r Indicates the UAV transmission power, Denotes the maximum transmission power of the UAV, and defines η sr Indicates the total hardware damage from the base station to the drone receiving end, subject to where κ sr Indicates the overall hardware damage level between the base station and the drone, meeting where κ s ,κ r Respectively represent the hardware damage suffered at the base station and the drone, η rn Indicates the total hardware damage from the drone to the user end, subject to where κ rn Indicates the overall hardware damage level between the drone and the user end, satisfying where κ r ,κ n They represent the hardware damage suffered by the drone and user n respectively; the tolerance is set to δ, and the 3D coordinates of the base station and the user are (x s ,y s ,0),(x n ,y n ,0),n=1,...,N; the initial position of the UAV is set to (x,y,H), where H represents the altitude of the UAV, and the channel coefficient from the base station to the UAV is defined as h sr , the channel coefficient from the UAV to the user terminal n is h rn ;Decoding order sequence satisfy where α k,n It is a binary variable; the number of iterations l, the calculation system downlink rate R sum , establish the system downlink and rate maximization problem P1: Constraint C1a indicates that the user power is non-negative; Constraint C1b indicates that the sum of the user powers is less than or equal to the UAV base station transmit power; Constraint C1c indicates the UAV transmit power limit; Constraint C1d indicates the base station transmit power limit; Constraint C1e indicates that the minimum rate of user n is higher than the quality of service threshold Constraint C1f represents user fairness; Constraint C1g indicates that user n cannot treat its own signal as interference; Constraint C1h represents the decoding order relationship between any two users k and n; Constraint C1i indicates that for any two users n and k, one user must be able to decode and eliminate the other user; Constraint C1j means that when a k,n =1, the distance between user n and the drone is greater than the distance between user k and the drone; In step 3), the initialization power is introduced into problem P1, and the UAV position optimization problem P2 is obtained. Now let H 2 +(xx k ) 2 +(yy k ) 2 =d rk ,H 2 +(xx n ) 2 +(yy n ) 2 =d rn , the specific form is as follows: Constraint C2a means that user n cannot regard its own signal as interference; Constraint C2b means the decoding order relationship between any two users k and n; Constraint C2c means that for any two users n and k, one user must be able to decode and eliminate the other user; Constraint C2d means that when a k,n = 1, the distance between user n and the drone is farther than the distance between user k and the drone. Constraint C2e indicates that the rate of any user n should be greater than its preset threshold. The step 3) is achieved by introducing variables R=(R1, ..., R n ),T=(t1,...,t n ) to replace the objective function and use the SCA technique to scale the constraint C2b. P2 can be transformed into a standard convex optimization problem P3, which is as follows: Constraint C3a means that user n cannot treat its own signal as interference; constraint C3b means that for any two users n and k, one user must be able to decode and eliminate the other user; constraint C3c means that the binary decision variable α k,n The value range constraint of C3d represents the value range of binary decision variable α k,n Scaling; Constraint C3e means that when a k,n = 1, the distance between user n and the drone is farther than the distance between user k and the drone; the constraint C3f means that for the variable T = (t1, ..., t n ) scaling constraint; constraint C3g represents the scaling constraint for the variable R = (R1, ..., R2); constraint C3h represents that the rate of any user n-end should be greater than its preset threshold; At this point, the interior point method can be used to solve the position of the drone; In step 4), the UAV position in step 3) is substituted into problem P1 to obtain the user power allocation problem P4, which is as follows: Constraint C4a indicates that the sum of user powers should not be greater than the transmit power; constraint C4b indicates user fairness; constraint C4c indicates that user power is non-negative; constraint C4d indicates that the n-terminal rate of any user should be greater than its preset threshold; by performing Lagrangian transformation on the objective function, an approximate variable Υ=(Υ1,...,Υ n ) replace the complex fraction part and introduce approximate variables according to the quadratic transformation regulations P4 can be transformed into a standard convex optimization problem P5, which has the following specific form: in By order Can be obtained: Similarly set up Can be obtained: P=(p1,...,p n ) represents the user power allocation, Finally, we have: Constraint C5a indicates that the sum of user powers should not exceed the transmit power; constraint C5b indicates user fairness; constraint C5c indicates that user power is non-negative; constraint C5d indicates that the n-terminal rate of any user should be greater than its preset threshold At this point, the interior point method can be used to solve the user power allocation optimization problem; In step 5), the UAV position obtained in step 3 and the public power distribution obtained in step 4 are substituted into the following formula: To calculate the downlink and rate for update.

2. The method for maximizing the sum rate of a UAV-assisted NOMA system under hardware damage conditions according to claim 1 is characterized in that: In step 6), repeat step 3) UAV position optimization, step 4) user power allocation optimization, and step 5) system downlink and rate update to iteratively update the UAV position and power. l Indicates that when the lth iteration, the value of the sum rate R is taken until R is satisfied l+1 -R l When <δ, it means that the difference between the l+1th sum rate and the lth sum rate is within the tolerance δ, and the convergence condition is reached; based on the proposed user power allocation method and UAV position optimization method, the final output is the UAV deployment position (x, y) under the maximum downlink sum rate, the SIC decoding sequence A, the user allocation power P, and the system sum rate R sum .

Images