IRS-UAV-based auxiliary NOMA-SWIPT system resource optimization allocation method

Abstract

The invention relates to a resource allocation method for a wireless information and power transfer system, in particular to a resource allocation method for a non-orthogonal multiple access (Non-Orthogonal Multiple Access, NOMA) and a wireless information and power transfer (SWIPT) system based on the assistance of an intelligent reflecting surface (IRS) and an unmanned aerial vehicle (UAV). According to the method, the sum rate of the system is maximized under the constraint conditions of UAV tracks, base station transmitting power distribution, IRS reflection coefficients and the minimum transmission rate and energy collection of Internet of Things (IoT) devices. The method comprises the following steps: firstly, establishing an IRS-UAV assisted multi-cluster NOMA-SWIPT communication system model, and proposing a multi-target optimization problem; for a non-convex optimization problem, based on an alternating iteration strategy, the non-convex optimization problem is decomposed into a plurality of single-target optimization sub-problems, and the single-target optimization sub-problems are solved step by step through methods such as continuous convex approximation. Compared with a UAV system which does not use an IRS, the method effectively improves the sum rate of the system, and has better practicability and feasibility.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communications and relates to a method for optimizing resource allocation of a wireless energy-carrying communication-non-orthogonal multiple access system assisted by an unmanned aerial vehicle equipped with an intelligent reflective surface. Background Art

[0002] With the rapid development of Internet of Things technology, wireless cellular networks are required to provide seamless connections for massive IoT devices distributed around the world. Therefore, how to use limited resources to provide seamless cellular connection services for such a large number of IoT devices will face huge technical challenges. Traditional orthogonal multiple access technology reduces multiple access interference by allocating different orthogonal time domain / frequency domain resources to different IoT devices, but the maximum number of IoT devices supported by this technology is strictly limited by the available orthogonal wireless resources. Non-Orthogonal Multiple Access (NOMA) technology utilizes the differences in IoT device channels, uses superposition coding processing at the transmitting end, and uses continuous interference cancellation technology at the receiving end to eliminate interference from other devices. It can provide services for multiple IoT devices on the same resource, and allows different IoT devices to share the same time and frequency resources based on power reuse technology to achieve improved spectrum efficiency and energy efficiency.

[0003] In future cellular networks, the access of massive IoT devices will lead to a sharp increase in energy consumption. IoT devices are usually powered by batteries with limited capacity, so there is an urgent need to find new energy sources to alleviate the shortage of battery power. The wireless energy-carrying communication (SWIPT) technology has two typical receiver structures based on time switching (TS) and power splitting (PS). The TS-based SWIPT technology requires the receiver to implement information decoding and energy collection functions in different time periods; while the PS-based SWIPT technology requires the receiver to split the received signal power according to a certain ratio for energy collection and information decoding respectively. SWIPT allows IoT devices to use the received signals to simultaneously realize information decoding and energy collection, which is expected to bring significant gains in terms of spectrum efficiency and energy efficiency, and is therefore more in line with the requirements of future cellular IoT green communications.

[0004] For example, the prior art with publication number CN 116015374 is a method for secure beamforming based on intelligent reflecting surface (IRS) assisted non-orthogonal multiple access energy-carrying network, which effectively improves system performance by deploying IRS. The amplitude and phase of the incident signal are adjusted by IRS, so that the reflected signal can be intentionally enhanced or weakened on different receivers, thereby realizing the reconstruction of the wireless channel and effectively improving the performance of the communication system. However, the coverage range of a single IRS is limited. In the case of massive IoT device connections, IoT devices that are far away from the base station cannot be well covered. In the case of massive IoT device connections, it is difficult to fully solve the problem of communication system performance by simply deploying a single IRS. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a method for resource allocation of a NOMA-SWIPT system based on IRS-Unmanned Aerial Vehicle (UAV) assistance. The UAV equipped with IRS has dynamic mobility and flexibility, and has the advantages of easy deployment and low cost. At the same time, IRS utilizes the amplitude and phase of the incident signal to reconstruct the wireless channel from the base station (BS) to the IoT device, thereby effectively improving the performance of the communication system. Therefore, the use of IRS-UAV assisted communication system can better expand the coverage of the BS, improve the signal reception quality of devices far away from the BS, and effectively improve the coverage and transmission efficiency of the communication system.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A resource allocation method for a NOMA-SWIPT system based on IRS-UAV assistance, the method comprising the following steps:

[0008] Step S1: Deploy a NOMA-SWIPT system with T NOMA clusters, where each NOMA cluster deploys T k An IoT device uses a UAV equipped with IRS to provide auxiliary communication for the system. The IoT device uses wireless energy communication technology to use the received signal for information decoding and energy collection.

[0009] Step S2: Solve the multivariable coupling optimization problem. First, initialize the BS's transmit beamforming vector w, the UAV's initial flight trajectory Q, the IRS's reflection coefficient matrix Θ, the IoT device's PS coefficient ρ, and the power allocation coefficient α, and set the maximum number of iterations N. max , convergence decision threshold ε, where the subscript max Indicates the maximum value;

[0010] Step S3: Using the given IRS reflection phase shift coefficient, the BS transmit beamforming vector, the power allocation coefficient of the IoT device, and the PS coefficient ratio, the trajectory of the UAV is solved and updated;

[0011] Step S4: Using the given UAV trajectory, the reflection phase shift coefficient of the IRS, the power allocation coefficient of the IoT device, and the PS coefficient ratio, the transmit beamforming vector of the BS is solved and updated;

[0012] Step S5: Using the given UAV trajectory, the BS's transmit beamforming vector, and the IRS's reflection phase shift coefficient, the power allocation coefficient and PS coefficient ratio of the IoT device are solved and updated;

[0013] Step S6: Using the given UAV trajectory, the BS transmit beamforming vector, the power allocation coefficient of the IoT device, and the PS coefficient ratio, the IRS reflection phase shift coefficient is solved and updated;

[0014] Step S7: Determine whether the system total sum rate update has converged, calculate the updated system sum rate, and if the absolute value of the difference between the updated sum rate and the previous sum rate is not greater than the sum rate decision threshold, determine that the system sum rate has converged, give the maximum sum rate value, and end the method; if the absolute value of the difference between the updated sum rate and the previous sum rate is greater than the sum rate decision threshold, save the newly calculated sum rate value as the latest sum rate, and update the parameters from step 2 until the system sum rate meets the convergence conditions and the required maximum sum rate is given.

[0015] Further, in step S1, the constructed NOMA-SWIPT system includes a base station equipped with N antennas, a smart reflective surface equipped with M reflective units, and T NOMA clusters, and the t-th NOMA cluster includes T k IoT devices.

[0016] The complex baseband signal sent by BS is

[0017]

[0018] in Indicates the first to the Tth k The sum of the terms; [n] represents the nth time slot in the UAV flight, w t [n] represents the beamforming vector of the t-th NOMA cluster in the n-th time slot, s t,k [n] represents the signal sent to the kth IoT device in the tth NOMA cluster. t,k represents the kth IoT device in the tth NOMA cluster, α t,k[n] is the power allocation coefficient of the kth IoT device in the tth NOMA cluster;

[0019] The signal received by the IoT device in the nth time slot is:

[0020]

[0021] The channel from BS to IoT device is denoted as The channel from BS to UAV-IRS is denoted as The channel from UAV-IRS to IoT device is denoted as in and represent complex matrices of dimensions M×N and M×K respectively, is the phase shift matrix of IRS, where β m [n]∈(0,1) and θ m [n]∈[0,2π] represents the amplitude and phase of each passive reflector unit, diag(·) represents a diagonal matrix, each diagonal element is the corresponding element ·, and j represents the imaginary unit; The mean is 0 and the variance is The additive Gaussian white noise of , ~ represents the distribution, express.

[0022] Furthermore, in step S2, the optimization problem of the method is formed as follows:

[0023] The received signal to interference plus noise ratio (SINR) of IoT devices can be expressed as

[0024]

[0025] in Gaussian white noise is introduced when decoding information, and |·| represents the absolute value.

[0026] In addition, a PS receiver is deployed at the IoT device, which divides the received signal into two parts through the PS structure, one part is used for information decoding, and the remaining part is used for energy collection. Therefore, the part of the signal received by the IoT device for energy collection can be expressed as:

[0027]

[0028] According to the above formula, the input power of the IoT device energy harvesting (EH) receiver in the nth time slot is expressed as:

[0029]

[0030] The EH of an IoT device can be expressed as:

[0031]

[0032] in represents the energy collected by the IoT device from the circuit in the nth time slot, With input power The function, parameter M t,k It represents the maximum harvesting power that the IoT device can receive when the EH circuit is saturated. Parameter a t,k and b t,k Represents hardware parameters related to the circuit.

[0033] Therefore, the information reachability rate of the kth IoT device in the tth NOMA cluster at the nth time slot can be expressed as

[0034]

[0035] Where log2(·) represents the logarithmic function with base 2;

[0036] Under the constraints of user minimum transmission rate, EH collection threshold, UAV flight, BS maximum transmission power and IRS reflection phase shift, the optimization problem P1 is constructed with the goal of maximizing the total rate of the system:

[0037]

[0038] C3:0≤ρ t,k [n]≤1

[0039] C4:0≤θ m [n]≤2π

[0040] C5:q I =q u [0], q F =q u [N]

[0041] C6:v[0]=v0,v[N]=v F

[0042]

[0043] C8:||v[n]||≤v max ,||v[n]||≤v min ,||a[n]||≤a max

[0044]

[0045] in It means solving x and y to maximize the value of f(x,y), P· represents the optimization problem number, is the minimum transmission rate that IoT devices need to meet. The superscript min indicates the minimum value. t,k represents the energy harvesting threshold of the IoT device, Q represents the optimized trajectory of the drone, and q I and q N denote the flight start and end points of the UAV, α and ρ denote the power allocation coefficient and PS coefficient ratio of the IoT device, v[n] denotes the flight speed of the UAV in the nth time slot, a[n] denotes the acceleration of the UAV flight, ||·|| denotes the norm, P max The optimization variables of this problem are the flight trajectory Q of the UAV, the transmit beamforming vector w of the BS, the power allocation coefficient α and the PS ratio ρ, and the reflection phase shift Θ of the IRS, where the objective function R sum represents the system and rate, the subscript sum represents the sum of the rates of all IoT devices, constraint C1 represents the achievable rate constraint for IoT users, constraint C2 represents that the energy collected by IoT users is greater than the power consumption, constraint C3 represents the energy collection PS coefficient constraint, and its value should be between 0 and 1, constraint C4 represents the IRS reflection coefficient constraint, constraints C5, C6 and C7 represent the trajectory constraints and flight speed constraints of UAV, and P in constraint C9 max Indicates the maximum transmit power of BS, C 10 Represents the power allocation constraint of IoT devices. Based on the alternating iterative strategy, the original optimization problem is decomposed into four sub-problems and solved separately.

[0046] Further, in step S3, the following optimization problem is solved to obtain the trajectory optimization position of the UAV:

[0047]

[0048] stC1,C2,C5,C6,C7,C8.

[0049] The constraint C1 in the optimization problem is a non-convex constraint. It is transformed into a concave function difference through equivalent form conversion and the auxiliary variable c is introduced. RU [n] Introduce auxiliary variable c RU [n] and c BR [n] represents the upper bound of the distance from UAV-IRS to the user and the upper bound of the distance from BS to UAV-IRS, respectively. The auxiliary variable l RU [n] and l BR[n] represents the lower bound of the distance between UAV-IRS and the user and the lower bound of the distance between BS and UAV-IRS respectively. The concave function difference form is transformed into a convex constraint by using the continuous convex approximation method. Similarly, other non-convex constraints are transformed by using continuous convex approximation and other methods, and the non-convex optimization problem P2 is transformed into a convex optimization problem for solution.

[0050] Further, in step S4, the following optimization problem is solved to obtain the transmit beamforming vector of the BS:

[0051]

[0052] stC1,C2,C9.

[0053] Introducing auxiliary variables And the auxiliary variable W t [n]Satisfy Rank(W t [n])=1. According to the above transformation, the constraint formula is processed, the objective function form is transformed, and the above non-convex constraint is transformed into a convex constraint form by using methods such as first-order Taylor expansion and continuous convex approximation, thereby converting the non-convex optimization problem P3 into a convex optimization problem for solution. Since the rank-one constraint is difficult to solve, it is relaxed by SDR. After removing the rank-one constraint, the problem is a standard convex optimization problem, which can be effectively solved by convex optimization tools to obtain the optimal solution W t [n], and then the optimal BS transmit beamforming vector can be obtained through eigenvalue decomposition.

[0054] Further, in step S5, the following optimization problem is solved to obtain the power allocation coefficient and PS ratio of the IoT device:

[0055]

[0056] stC1,C2,C3,C 10 .

[0057] The optimization problem P4 uses mathematical transformation to transform the constraint formula through the above-solved UAV flight trajectory position, the BS's transmit beamforming vector and the given IRS transmit phase shift coefficient. After the non-convex constraints are converted into convex constraints through methods such as Taylor expansion and continuous convex approximation, the optimization problem is a standard convex optimization problem, which can solve the optimal power allocation coefficient and PS ratio of the IoT device.

[0058] Further, in step S6, the following optimization problem is solved to obtain the optimal phase shift of the IRS reflection coefficient:

[0059]

[0060] stC1,C2,C4.

[0061] Optimization problem P5 optimizes the reflection phase shift of the IRS through the above-solved UAV flight trajectory position, BS's transmit beamforming vector, IoT device's power allocation coefficient and PS ratio. The continuous convex approximation method and Taylor expansion method are used to transform the non-convex constraints into convex constraints. The semi-positive definite relaxation technique is then used to relax the difficult-to-solve constraints. The optimization problem is transformed into a standard convex optimization problem. Finally, the Gaussian randomization method is used to obtain a solution that satisfies the rank-one condition.

[0062] The beneficial results of the present invention are:

[0063] The present invention is a resource optimization allocation method for a NOMA-SWIPT system assisted by IRS-UAV. By optimizing the flight trajectory of UAV, the transmit beamforming vector of BS, the power allocation coefficient of IoT device, PS coefficient ratio and the reflection beamforming vector of IRS, the optimization problem is a non-convex optimization problem with multivariable coupling. The optimization problem is decomposed into multiple sub-problems through mathematical form transformation, continuous convex approximation, Taylor expansion method, semi-positive relaxation method and alternating iterative optimization. For a single optimization problem, the non-convex constraint is converted into a convex constraint, and the non-convex optimization problem can be converted into a standard convex optimization problem, thereby obtaining the optimal solution of multiple variable values. By using the UAV equipped with IRS to provide auxiliary communication for the communication system, its flexible mobility can be used to effectively improve the communication quality of IoT devices. At the same time, the use of wireless energy-carrying communication can solve the problem of insufficient energy for IoT devices while receiving signals. A method based on alternating iterative optimization strategy is adopted to solve multiple optimization variables. The proposed scheme provides a potential solution for seamless cellular connection services of massive IoT devices in the future.

[0064] The present invention uses IRS-UAV to assist the NOMA-SWIPT system, and transforms the system's beamforming vector, UAV trajectory optimization, IRS reflection phase shift coefficient, power allocation coefficient of IoT devices, and PS ratio into standard convex optimization problems and solves them respectively. On the basis of efficient clustering of IoT devices, NOMA technology is used to achieve massive IoT device connection, effectively expanding the coverage of the BS, improving the communication transmission quality for IoT devices far away from the BS, effectively improving the total rate of the system, alleviating the problem of insufficient battery function of IoT devices, and bringing significant performance gains to the communication system, which is more in line with the requirements of future cellular IoT green communications.

[0065] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be described in detail below in conjunction with the accompanying drawings, wherein:

[0067] Figure 1 Schematic diagram of the IRS-UAV-assisted NOMA-SWIPT system model;

[0068] Figure 2 The overall flow chart of the system resource optimization allocation method provided by the present invention; DETAILED DESCRIPTION

[0069] The following describes the embodiments of the present invention by specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner, and the following embodiments and features in the embodiments can be combined with each other without conflict.

[0070] Among them, the drawings are only used for illustrative explanations, and they only represent schematic diagrams rather than actual pictures, and should not be understood as limitations on the present invention. In order to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of actual products. For those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings may be omitted.

[0071] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "front", "rear", etc. indicate the orientation or position relationship, they are based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation. Therefore, the terms describing the position relationship in the drawings are only used for illustrative purposes and cannot be understood as limiting the present invention. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0072] See also Figure 1-2 ,A resource allocation method for NOMA-SWIPT system based on IRS-UAV assistance.

[0073] In the scenario of massive IoT device connections under 5G, the use of NOMA technology can effectively improve the resource utilization of the system. At the same time, the use of IRS-UAV to assist the NOMA-SWIPT system can effectively improve the communication quality of IoT devices. The wireless communication channel is reconfigured using IRS to provide additional receiving paths for IoT devices, effectively enhancing the receiving signal of the device, and realizing seamless cellular connection services in the current scenario.

[0074] At present, the IRS-UAV assisted NOMA-SWIPT system is still in its preliminary stage. Most of the research focuses on the IRS assisted NOMA-SWIPT system, without considering the limited coverage of a single IRS. In addition, there are few studies on the use of IRS-UAV assisted NOMA-SWIPT system in the scenario of massive IoT devices connected. Therefore, the research on the system resource allocation method of the IRS-UAV assisted NOMA-SWIPT system needs to be further deepened.

[0075] In view of this, the purpose of the present invention is to provide a method for resource allocation of a NOMA-SWIPT system based on IRS-UAV assistance. The UAV equipped with IRS has dynamic mobility and flexibility, and has the advantages of easy deployment and low cost. At the same time, IRS reconstructs the wireless channel from BS to IoT devices by adjusting the amplitude and phase of the incident signal, thereby effectively improving the performance of the communication system. Therefore, the use of IRS-UAV assisted communication system can better expand the coverage of the BS, improve the signal reception quality of devices far away from the BS, and effectively improve the coverage and transmission efficiency of the communication system.

[0076] In order to achieve the above object, the present invention provides the following technical solutions:

[0077] A resource allocation method for a NOMA-SWIPT system based on IRS-UAV assistance, the method comprising the following steps:

[0078] Step S1: Deploy a NOMA-SWIPT system with T NOMA clusters, where each NOMA cluster deploys T k An IoT device uses a UAV equipped with IRS to provide auxiliary communication for the system. The IoT device uses wireless energy communication technology to use the received signal for information decoding and energy collection.

[0079] Step S2: Solve the multivariable coupling optimization problem. First, initialize the BS's transmit beamforming vector w, the UAV's initial flight trajectory Q, the IRS's reflection coefficient matrix Θ, the IoT device's PS coefficient ρ, and the power allocation coefficient α, and set the maximum number of iterations N. max , convergence decision threshold ε, where the subscript max Indicates the maximum value;

[0080] Step S3: Using the given IRS reflection phase shift coefficient, the BS transmit beamforming vector, the power allocation coefficient of the IoT device, and the PS coefficient ratio, the trajectory of the UAV is solved and updated;

[0081] Step S4: Using the given UAV trajectory, the reflection phase shift coefficient of the IRS, the power allocation coefficient of the IoT device, and the PS coefficient ratio, the transmit beamforming vector of the BS is solved and updated;

[0082] Step S5: Using the given UAV trajectory, the BS's transmit beamforming vector, and the IRS's reflection phase shift coefficient, the power allocation coefficient and PS coefficient ratio of the IoT device are solved and updated;

[0083] Step S6: Using the given UAV trajectory, the BS transmit beamforming vector, the power allocation coefficient of the IoT device, and the PS coefficient ratio, the IRS reflection phase shift coefficient is solved and updated;

[0084] Step S7: Determine whether the system total sum rate update has converged, calculate the updated system sum rate, and if the absolute value of the difference between the updated sum rate and the previous sum rate is not greater than the sum rate decision threshold, determine that the system sum rate has converged, give the maximum sum rate value, and end the method; if the absolute value of the difference between the updated sum rate and the previous sum rate is greater than the sum rate decision threshold, save the newly calculated sum rate value as the latest sum rate, and update the parameters from step 2 until the system sum rate meets the convergence conditions and the required maximum sum rate is given.

[0085] Further, in step S1, the constructed NOMA-SWIPT system includes a base station equipped with N antennas, a smart reflective surface equipped with M reflective units, and T NOMA clusters, and the t-th NOMA cluster includes T k IoT devices.

[0086] The complex baseband signal sent by the BS is

[0087]

[0088] in Indicates the first to the Tth k The sum of the terms; [n] represents the nth time slot in the UAV flight, wt [n] represents the beamforming vector of the t-th NOMA cluster in the n-th time slot, s t,k [n] represents the signal sent to the kth IoT device in the tth NOMA cluster. t,k represents the kth IoT device in the tth NOMA cluster, α t,k [n] is the power allocation coefficient of the kth IoT device in the tth NOMA cluster;

[0089] The signal received by the IoT device in the nth time slot is:

[0090]

[0091] The channel from BS to IoT device is denoted as The channel from BS to UAV-IRS is denoted as The channel from UAV-IRS to IoT device is denoted as in and represent complex matrices of dimensions M×N and M×K respectively, is the phase shift matrix of IRS, where β m [n]∈(0,1) and θ m [n]∈[0,2π] represents the amplitude and phase of each passive reflector unit, diag(·) represents a diagonal matrix, each diagonal element is the corresponding element ·, and j represents the imaginary unit; The mean is 0 and the variance is The additive Gaussian white noise of , ~ represents the distribution, represents a complex Gaussian distribution.

[0092] Furthermore, in step S2, the optimization problem of the method is formed as follows:

[0093] The received signal-to-interference-noise ratio of IoT devices can be expressed as

[0094]

[0095] in The mean is 0 and the variance is where i represents other interfering devices in the cluster and j represents the beams corresponding to other clusters.

[0096] In addition, a PS receiver is deployed at the IoT device, which divides the received signal into two parts through the PS structure, one part is used for information decoding, and the remaining part is used for energy collection. Therefore, the part of the signal received by the IoT device for energy collection can be expressed as:

[0097]

[0098] Therefore, the information reachability rate of the kth IoT device in the tth NOMA cluster at the nth time slot can be expressed as

[0099] R t,k [n] = log2(1 + SINR t,k [n])

[0100] Where log2(·) represents the logarithmic function with base 2;

[0101] Under the constraints of user minimum transmission rate, energy collection threshold, UAV flight, BS maximum transmission power, and IRS reflection phase shift, the optimization problem P1 is constructed with the goal of maximizing the total rate of the system:

[0102]

[0103] C3:0≤ρ t,k [n]≤1

[0104] C4:0≤θ m [n]≤2π

[0105] C5:q I =q u [0], q F =q u [N]

[0106] C6:v[0]=v0,v[N]=v F

[0107]

[0108] C8:||v[n]||≤v max ,||v[n]||≥v min ,||a[n]||≤a max

[0109]

[0110] in It means solving x and y to maximize the value of f(x,y), P· represents the optimization problem number, is the minimum transmission rate that IoT devices need to meet. The superscript min indicates the minimum value. t,k represents the energy harvesting threshold of the IoT device, Q represents the optimized trajectory of the drone, and q I and q Ndenote the flight start and end points of the UAV, α and ρ denote the power allocation coefficient and PS coefficient ratio of the IoT device, v[n] denotes the flight speed of the UAV in the nth time slot, a[n] denotes the acceleration of the UAV flight, ||·|| denotes the norm, P max The optimization variables of this problem are the flight trajectory Q of the UAV, the transmit beamforming vector w of the BS, the power allocation coefficient α and the PS ratio ρ, and the reflection phase shift Θ of the IRS, where the objective function R sum represents the system and rate, the subscript sum represents the sum of the rates of all IoT devices, constraint C1 represents the achievable rate constraint for IoT users, constraint C2 represents that the energy collected by IoT users is greater than the power consumption, constraint C3 represents the energy collection PS coefficient constraint, and its value should be between 0 and 1, constraint C4 represents the IRS reflection coefficient constraint, constraints C5, C6 and C7 represent the trajectory constraints and flight speed constraints of UAV, and P in constraint C9 max Indicates the maximum transmit power of BS, C 10 Represents the power allocation constraint of IoT devices. Based on the alternating iterative strategy, the original optimization problem is decomposed into four sub-problems and solved separately.

[0111] Considering that the above optimization problem is a complex non-convex optimization problem with multi-variable coupling and contains multiple non-convex optimization constraints, it is difficult to directly solve the complex optimization problem. Therefore, a method based on alternating iterative optimization algorithm is given below to solve the above optimization problem, which decomposes the complex optimization problem into four optimization problems and solves them separately.

[0112] Further, in step S3, the following optimization problem is solved to obtain the trajectory optimization position of the UAV:

[0113]

[0114] stC1,C2,C5,C6,C7,C8.

[0115] First, the objective function can be equivalently transformed into the following form:

[0116]

[0117] By introducing the auxiliary variable l tk [n] and u tk [n], the above formula can be transformed into the following form:

[0118]

[0119] The auxiliary variables satisfy the following constraints

[0120]

[0121] The effective channel gain of IoT devices is mathematically transformed into the following form:

[0122]

[0123] Some of the items can be expressed as

[0124]

[0125] Where Re(·) represents the operation of taking the real part, (·) T Represents the transpose operation on its elements.

[0126]

[0127] in(·) H It means to perform conjugate transpose operation on the elements.

[0128] Therefore, the effective channel gain can be converted into the following form

[0129]

[0130] Some of the items can be expressed as

[0131]

[0132] Similarly, there are the following forms of conversion

[0133]

[0134] In order to process the above formula, an auxiliary variable c is introduced. RU [n] and c BR [n] represents the upper bound of the distance from UAV-IRS to the user and the upper bound of the distance from BS to UAV-IRS, respectively. The auxiliary variable l RU [n] and l BR [n] represents the lower bound of the distance between UAV-IRS and user and the lower bound of the distance between BS and UAV-IRS, respectively.

[0135] The auxiliary variable satisfies the following relationship

[0136] (c RU [n]) 2 ≥||q u [n]-w t,k || 2 , (c BR [n]) 2 ≥||q u [n]-wB || 2

[0137] ||q u [n]-w t,k || 2 ≥(l RU [n]) 2 ,||q u [n]-w B || 2 ≥(l BR [n]) 2

[0138] Using continuous convex approximation and other methods on the above variables, it can be equivalently transformed into the following form

[0139]

[0140]

[0141] in Indicates q u [n] The feasible point at the mth iteration.

[0142] For the convenience of subsequent expression, the above formula is expressed as C 11-1 , C 11-2 , C 11-3 , C 11-4 .

[0143] make and Then there are as follows:

[0144]

[0145] Therefore, the above constraint formula can be derived according to the above formula and converted into the following form:

[0146]

[0147] For the above formula, the left formula needs to be converted into a concave function and processed. The processed constraint formula is marked as C 12-1

[0148] Similarly, the following constraints can be transformed

[0149]

[0150] Similar to the above processing method, the left side of the inequality needs to be transformed into a convex function form, and the right side of the inequality needs to be transformed into a concave function form, which is as follows. For the convenience of subsequent expression, the processed constraint is marked as C 12-2 .

[0151]

[0152] in Indicates u tk [n] The feasible point at the mth iteration.

[0153] Constraint C1 can also be transformed into the following form, which is marked as C 1-1 .

[0154]

[0155] For constraint C2, use the above method to convert the form

[0156]

[0157] The effective channel gain of the left formula is transformed into a concave function form by the above mathematical transformation, and the non-convex constraint can be transformed into a convex constraint, and the processed form can be transformed into C 2-1 .

[0158] The constraint C8 is mathematically transformed as follows and marked as C 8-1 .

[0159]

[0160] Where v0[n] represents the feasible value of v[n] at the mth iteration.

[0161] After the above constraint transformation, the optimization problem can be transformed into P2'

[0162]

[0163] sC 1-1 ,C 2-1 ,C5,C6,C7,C 8-1 ,C 11-1 ,C 11-2 ,C 11-3 ,C 11-4 ,C 12-1 ,C 12-2 .

[0164] The optimal trajectory position of the UAV can be obtained by solving the problem using the CVX toolbox.

[0165] Further, in step S4, the following optimization problem is solved to obtain the transmit beamforming vector of the BS:

[0166]

[0167] stC1,C2,C9.

[0168] The optimization problem is non-convex. Since the beamforming vector and the user's power allocation coefficient exist in a coupled form in most cases, it is necessary to deal with this non-convex problem. And the auxiliary variable W t [n]Satisfy Rank(W t [n])=1, where Rank(·) represents the rank of the matrix. Then we have in

[0169] According to the above formula, we have the following form

[0170]

[0171] where Tr(·) represents the trace of the matrix.

[0172] Therefore, the objective function can be transformed into the following form

[0173]

[0174] Further introduce auxiliary variable m t [n]=[m t,1 [n],m t ,2 [n]] T , so the auxiliary variables satisfy the following constraints:

[0175]

[0176] Therefore, the objective function can be equivalently transformed into the following form:

[0177]

[0178] It is necessary to continue processing the inequality after introducing the auxiliary variable, and introduce the auxiliary variable r t [n]=[r t,1 [n],r t,2 [n]] T Then we have:

[0179] log2(r t,1 [n])≥m t,1 [n]

[0180] Continue to transform the above formula and use a Taylor expansion to get the following form

[0181]

[0182] log2(r t,2 [n])≤m t,2 [n]

[0183]

[0184] For the convenience of subsequent description, the above are marked as C 13-1 ,C 13-2 ,C 13-3 ,C 13-4 .

[0185] Based on the above, the constraint C1 can be expressed as follows, which is marked as C 1-2 .

[0186]

[0187] At the same time, the constraint C2 can be expressed as follows, which is marked as C 2-2 .

[0188]

[0189] in

[0190] Similarly, the constraint C9 is expressed as follows and marked as C 9-1

[0191]

[0192] Therefore, the optimization problem P3 can be transformed into the following form:

[0193]

[0194] sC 1-2 ,C 2-2 ,C 9-1 ,C 13-1 ,C 13-2 ,C 13-3 ,C 13-4 .

[0195]

[0196] C 14-2 :Rank(W t )≤1

[0197] Since the constraint C 14-2 It is difficult to solve, so SDR is used to relax it. After removing the rank-one constraint, this problem is a standard convex optimization problem, which can be effectively solved by the convex optimization toolbox to obtain the optimal solution. Then the optimal transmit beamforming vector can be obtained by eigenvalue decomposition

[0198] Further, in step S5, the following optimization problem is solved to obtain the power allocation coefficient and PS ratio of the IoT device:

[0199]

[0200] stC1,C2,C3,C 10 .

[0201] This optimization problem is a non-convex optimization problem. Auxiliary variables are introduced. The equivalent channel gain of the user can be expressed as:

[0202]

[0203] Introduce auxiliary variable p t [n]=[p t,1 [n],p t,2 [n]] T , satisfying the following relationship

[0204]

[0205] For the convenience of subsequent description, the above are marked as C 15-1 ,C 15-2 .

[0206] Due to the form of the PS ratio coefficient, the above constraints are non-convex constraints, so an auxiliary variable p is introduced a [n], the change amount satisfies the following constraint relationship, which is marked as C 16-1 .

[0207]

[0208] Constraint C 16-1 It is still a non-convex constraint. Using the first-order Taylor expansion, we can get the following form, which is marked as C 16-2 .

[0209]

[0210] in Represents ρ t,k [n] The feasible value at the nth iteration.

[0211] According to the constraint C 16-1 The constraint C 15-1 Convert it into the following form and mark it as C 15-3 .

[0212]

[0213] It is necessary to continue processing the inequality after the introduction of auxiliary variables and introduce auxiliary variables g t [n]=[g t,1 [n],g t,2 [n]] T Then we have:

[0214] log2(g t,1 [n])≥p t,1 [n]

[0215] log2(g t,2 [n])≤p t,2 [n]

[0216] It is necessary to use continuous convex approximation to convert the above formula into the following form, which are marked as C 17-1 ,C 17-2 .

[0217] log2(g t,1 [n])≥p t,1 [n]

[0218]

[0219] According to the above analysis, the objective function can be equivalently transformed into the following form:

[0220]

[0221] Similarly, constraint C1 can be transformed into the following equivalent form, which is marked as C 1-3 .

[0222]

[0223] Expand constraint C2 into the following form and mark it as C 2-3 .

[0224]

[0225] in

[0226] Since IoT devices need to meet both the minimum transmission rate constraint and the minimum energy collection constraint, they have the following form

[0227]

[0228] Based on the above, the constraint C3 is rewritten as follows and marked as C 3-1 .

[0229]

[0230] According to the above transformation, the optimization problem P4 can be transformed into the following form:

[0231]

[0232] sC 1-3 ,C 2-3 ,C 3-1 ,C 10 ,C 15-2 ,C 15-3 ,C 16-2 ,C 17-1 ,C 17-2 .

[0233] The optimization problem P4' is transformed into a standard convex optimization problem, which can be solved using the CVX toolbox to obtain the optimal power allocation coefficient and PS coefficient ratio of the IoT device.

[0234] Further, in step S6, the following optimization problem is solved to obtain the optimal phase shift of the IRS reflection coefficient:

[0235]

[0236] stC1,C2,C4.

[0237] The optimization problem P5 is a non-convex optimization problem. First, the non-convex constraint C1 is processed, where the effective channel gain of the IoT device is transformed. It can be equivalently converted into the following form:

[0238]

[0239] in

[0240] So define r t [n] = Ψ t [n]w t [n], R t [n] = r t [n]r t H [n], In addition,

[0241]

[0242] Because the mathematical transformation can be obtained as follows

[0243]

[0244] According to the above formula, the effective channel gain can be converted into the following form

[0245]

[0246] Introducing auxiliary variables Its satisfaction

[0247] Based on the above derivation, the constraint C1 can be transformed into the following mathematical form:

[0248]

[0249] Perform mathematical transformation on the left side of the above equation and convert it into the equivalent form of concave function difference.

[0250]

[0251]

[0252] Therefore, the constraint can be transformed into the following form

[0253]

[0254] Since the constraint C1 is still in a non-convex form, it is in the form of a convex function difference. j,i [n] The upper bound is obtained using the continuous convex approximation method, so by applying the first-order Taylor expansion, we obtain:

[0255]

[0256] Where V (m) [n] represents the feasible value of V[n] at the mth iteration, and some terms can be expressed as

[0257]

[0258] Therefore, according to the above derivation, the constraint C1 can be transformed into the following form, which is marked as C 1-4 .

[0259]

[0260] where g j,i [n](V (m) [n]) The value of at the mth iteration is obtained by applying a first-order Taylor expansion.

[0261] Similarly, the objective function can be approximated as follows, where the superscript ub indicates the upper limit of the formula.

[0262]

[0263] Mathematically transform the constraint C2, and we can get the following form. For the convenience of subsequent expression, we will mark it as C 2-4 .

[0264]

[0265] in

[0266] Based on the above derivation, the phase shift optimization subproblem P5 can be expressed as follows:

[0267]

[0268] sC 1-4 ,C 2-4 .

[0269]

[0270] The optimization problem P5' is a standard convex optimization problem, which can be solved using a convex optimization solver (such as CVX). In general, solving the problem P5' cannot obtain a rank-one solution, i.e. Therefore, Gaussian randomization (SDR) is used to achieve the solution of problem P5'.

[0271] like So The eigenvalue decomposition of can be expressed as:

[0272]

[0273] where T = [v1, v2, …, v M+1 ] represents the identity matrix of eigenvectors, Σ=diag(λ1,λ2,...,λ M+1 ) represents the diagonal matrix of eigenvalues. Then two independent zero-mean normally distributed random vectors β, χ and the covariance matrix are randomly generated.

[0274] Based on the obtained Gaussian random variable x t , we can obtain the suboptimal solution to problem (P1.1), which is expressed as follows:

[0275]

[0276] According to the above content, the present invention is a resource optimization allocation method for the NOMA-SWIPT system assisted by IRS-UAV. By designing the transmission beam vector of the BS, the flight trajectory of the UAV, the reflection phase shift coefficient matrix of the IRS, the power allocation coefficient of the IoT and the PS coefficient ratio, a complex optimization problem with multivariable coupling is constructed. First, the original non-convex optimization problem is decomposed into four sub-problems based on the alternating iterative optimization algorithm, and then multiple non-convex constraints are converted into convex constraints by using continuous convex approximation, Taylor expansion and semi-positive relaxation. The non-convex optimization problem is converted into a standard convex optimization problem, and the CVX toolbox is used to solve it to obtain the optimal values ​​of multiple optimization variables. On the basis of efficient clustering of IoT devices, NOMA technology is used to realize the connection of massive IoT devices, effectively expanding the coverage of the BS, improving the communication transmission quality for IoT devices far away from the BS, effectively improving the total rate of the system, alleviating the problem of insufficient battery function of IoT devices, and bringing significant performance gains to the communication system, which is more in line with the requirements of future cellular IoT green communication.

[0277] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the purpose and scope of the technical solution, which should be included in the scope of the claims of the present invention.

Claims

1. A method for resource allocation based on intelligent reflecting surface (IRS)-unmanned aerial vehicle (UAV)-assisted non-orthogonal multiple access (NOMA)-simultaneous wireless information and power transfer (SWIPT) system, characterized in that: The following steps are involved: Step S1: Establish a sum rate maximization model of an IRS-UAV-assisted NOMA-SWIPT system; Step S2: Solve the optimization problem by first initializing the UAV position, IRS reflection phase shift, base station transmit beamforming vector, power allocation coefficient of Internet of things (IoT), power splitting coefficient (PS) ratio, and sum rate convergence threshold; Step S3: Using the given IRS reflection phase shift, the base station (BS) transmit beamforming vector, the power allocation coefficient of the IoT device, and the PS ratio, the UAV trajectory is solved and updated; Step S4: Using the solved UAV trajectory and the given IRS reflection phase shift, IoT power allocation coefficient, and PS split coefficient, solve the BS transmit beamforming vector and update it; Step S5: using the solved UAV trajectory, the BS transmit beamforming vector and the given IRS phase shift, solve the power allocation coefficient and PS coefficient ratio of the IoT device and update them; Step S6: using the solved UAV trajectory, the transmit beamforming vector of the BS, the power allocation coefficient of the IoT device, and the PS coefficient ratio, solve the reflection phase shift of the IRS and update it; Step S7: Determine whether the sum rate update has converged. Calculate the updated system sum rate. If the absolute value of the difference between the updated sum rate and the previous sum rate is not greater than the sum rate decision threshold, determine that the system sum rate has converged, give the maximum sum rate value, and end the method. If the absolute value of the difference between the updated sum rate and the previous sum rate is greater than the sum rate decision threshold, save the newly calculated sum rate value as the latest sum rate, and update the parameters from step 2 until the system sum rate meets the convergence conditions and the required maximum sum rate is given.

2. The method for maximizing the sum rate based on the IRS-UAV assisted NOMA-SWIPT system according to claim 1, characterized in that: In step 1, the IRS-UAV-assisted NOMA-SWIPT system includes a base station equipped with N antennas, a smart reflective surface equipped with M reflective units, and T NOMA clusters, the tth NOMA cluster including T k IoT devices. The complex baseband signal sent by BS is in Indicates the first to the Tth k The sum of the terms; [n] represents the nth time slot in the UAV flight, w t [n] represents the beamforming vector of the t-th NOMA cluster in the n-th time slot, s t,k [n] represents the signal sent to the kth IoT device in the tth NOMA cluster. t,k represents the kth IoT device in the tth NOMA cluster, α t,k [n] is the power allocation coefficient of the kth IoT device in the tth NOMA cluster; The signal received by the IoT device in the nth time slot is: The channel from BS to IoT device is denoted as The channel from BS to UAV-IRS is denoted as The channel from UAV-IRS to IoT device is denoted as in and represent complex matrices of dimensions M×N and M×K respectively, is the phase shift matrix of IRS, where β m [n]∈(0,1) and θ m [n]∈[0,2π] represents the amplitude and phase of each passive reflector unit, diag(·) represents a diagonal matrix, each diagonal element is the corresponding element ·, and j represents the imaginary unit; The mean is 0 and the variance is The additive Gaussian white noise of , ~ represents the distribution, represents a complex Gaussian distribution. The composite channel gain from BS to IoT users in the nth time slot with the assistance of UAV-IRS can be expressed as The part of the received signal of the IoT device used for information decoding can be expressed as in The mean is 0 and the variance is Additive white Gaussian noise.

3. The IRS-UAV-assisted NOMA-SWIPT system resource allocation method according to claim 2, characterized in that: In step 2, the optimization problem of the method is formulated as follows: The received signal to interference plus noise ratio (SINR) of IoT devices can be expressed as in Gaussian white noise introduced during information decoding, |·| represents the absolute value, ρ t,k [n] represents the PS coefficient ratio of the IoT device, and the subscript j represents the beamforming vectors corresponding to the clusters other than the cluster corresponding to the IoT device. In addition, a PS receiver is deployed at the IoT device, which divides the received signal into two parts through the PS structure, one part is used for information decoding, and the remaining part is used for energy collection. Therefore, the part of the signal received by the IoT device for energy collection can be expressed as: According to the above formula, the input power of the IoT device energy harvesting (EH) receiver in the nth time slot is expressed as: The EH of an IoT device can be expressed as: in represents the energy collected by the IoT device from the circuit in the nth time slot, With input power The function, parameter M t,k It represents the maximum harvesting power that the IoT device can receive when the EH circuit is saturated. Parameter a t,k and b t,k Represents hardware parameters related to the circuit. Therefore, the information reachability rate of the kth IoT device in the tth NOMA cluster at the nth time slot can be expressed as Where log2(·) represents the logarithmic function with base 2; Under the constraints of user minimum transmission rate, EH threshold, UAV flight, BS maximum transmission power and IRS reflection phase shift, the optimization problem P1 is constructed with the goal of maximizing the total rate of the system: in It means solving x and y to maximize the value of f(x,y), P· represents the optimization problem number, is the minimum transmission rate that IoT devices need to meet. The superscript min indicates the minimum value. t,k Indicates the energy harvesting threshold of IoT devices, subscript t,k represents the kth IoT device of the corresponding tth cluster, Q represents the optimized trajectory of the UAV, α and ρ represent the power allocation coefficient and PS coefficient ratio of the IoT device, v[n] represents the flight speed of the UAV in the nth time slot, a[n] represents the acceleration of the UAV flight, ||·|| represents the norm, P max The optimization variables of this problem are the flight trajectory Q of the UAV, the transmit beamforming vector w of the BS, the power allocation coefficient α and the PS ratio ρ, and the reflection phase shift Θ of the IRS, where the objective function R sum represents the system and rate, the subscript sum represents the sum of the rates of all IoT devices, constraint C1 represents the achievable rate constraint for IoT users, constraint C2 represents that the energy collected by IoT users is greater than the power consumption, constraint C3 represents the energy collection PS coefficient constraint, and its value should be between 0 and 1, constraint C4 represents the IRS reflection coefficient constraint, constraints C5, C6 and C7 represent the trajectory constraints and flight speed constraints of UAV, and P in constraint C9 max Indicates the maximum transmit power of BS, C 10 Represents the power allocation constraint of IoT devices. Based on the alternating iterative strategy, the original optimization problem is decomposed into four sub-problems and solved separately.

4. The method for maximizing the sum rate based on the IRS-UAV assisted NOMA-SWIPT system according to claim 3 is characterized in that: In step 3, the UAV trajectory is solved and updated using the given IRS reflection phase shift, the BS transmit beamforming vector, the power allocation coefficient of the IoT device, and the PS ratio, which specifically includes: Fix Θ, w, ρ and α, optimize Q, and the sub-problem P2 is: The optimization problem P2 uses methods such as continuous convex approximation to effectively transform non-convex constraints into convex constraints, and transforms the original non-convex optimization problem into a convex optimization problem for solution.

5. According to claim 4, a method for resource allocation based on IRS-UAV assisted NOMA-SWIPT system, wherein in step 4, the transmit beamforming vector of the BS is solved and updated by using the solved UAV trajectory and the given IRS reflection phase shift, the power allocation coefficient of IoT and the PS division coefficient, specifically comprising: The optimization problem P3 is solved by converting the objective function into a concave function difference, introducing auxiliary variables, and using methods such as first-order Taylor expansion on the converted objective function to convert the non-convex optimization problem into a convex optimization problem.

6. According to claim 5, a method for resource allocation based on IRS-UAV assisted NOMA-SWIPT system, wherein in step 5, the power allocation coefficient and PS coefficient ratio of the IoT device are solved by using the solved UAV trajectory, the transmit beamforming vector of the BS and the given IRS phase shift, and are updated, specifically comprising: The optimization problem P4 converts the non-convex objective function into the form of concave function difference, and applies continuous convex approximation and Taylor expansion to convert the non-convex optimization problem into a convex optimization problem for solution.

7. According to claim 6, a method for resource allocation based on IRS-UAV assisted NOMA-SWIPT system, wherein in step 6, the reflection phase shift of IRS is solved and updated by using the solved UAV trajectory, the transmit beamforming vector of BS, the power allocation coefficient of IoT device and the PS coefficient ratio, which specifically includes: The optimization problem P5 uses continuous convex approximation and other methods to transform the non-convex optimization problem into a convex optimization problem, and finally uses the Gaussian randomization method to obtain a solution that satisfies the rank-one condition.

Images