A Method for Jointly Optimizing User Scheduling, Power Allocation, and UAV Trajectory for NOMA Communication Coverage

Abstract

The present invention discloses a method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage. According to the free space path loss model of communication between the UAV and users, the channel gain between each user and the UAV at each moment is obtained. After sorting the users according to the channel gain, the user scheduling and power allocation of the UAV under a fixed trajectory are obtained through dynamic programming. Based on this, the trajectory problem is transformed into a solvable convex optimization problem, and the trajectory under fixed user scheduling and power allocation is obtained. The user scheduling and power allocation of the UAV under the fixed trajectory and the trajectory under the fixed user scheduling and power allocation are iteratively repeated to obtain the optimal user scheduling, optimal power allocation, and the optimal trajectory of the UAV within the period, so as to maximize the energy efficiency of the UAV's communication coverage in the area under the condition of satisfying the constraints between the UAV and users.

Description

Technical Field

[0001] The present invention belongs to the technical field of UAV trajectory planning, and specifically relates to a method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage. Background Technique

[0002] Due to the advantages of small size, strong mobility, convenient deployment, good channel conditions, low cost, etc. of UAVs, UAVs are increasingly widely used in military and civilian fields. For example, collaborative reconnaissance and reasoning, disaster detection, mobile base station communication coverage, data relay services, etc. With the development of various derivative solutions for assisting communication, the role of UAVs in assisting communication in communication systems is becoming more and more obvious. Whether in the current cellular network or in the future 6G wireless communication network, UAVs will play an indispensable role.

[0003] In the scenario of UAV-assisted communication coverage, in order to provide better service quality for ground users, it is particularly important to plan the placement position of the UAV or the trajectory information of the UAV. At the same time, due to the limitations of on-board resources brought about by the small size of UAVs, such as computing power, UAV energy, etc., considering how to provide better service quality for users under the limited resources of UAVs has also become a huge challenge.

[0004] In existing research, many studies and attempts have been made on UAV-assisted cellular wireless networks, but these attempts mainly focus on technologies based on orthogonal multiple access, such as time division multiple access and frequency division multiple access technologies, to maximize the maximum capacity or max-min rate of the entire communication system. However, due to the limited energy of UAVs, how to maximize the system capacity in limited energy is particularly crucial. Technologies based on non-orthogonal multiple access can greatly improve the capacity of the entire system without additional requirements and constraints on UAVs. Therefore, in a UAV wireless network based on non-orthogonal multiple access technology, it is very important to maximize the system capacity per unit energy by jointly considering UAV power allocation and trajectory optimization while ensuring the communication needs of ground users. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage in view of the above-mentioned deficiencies of the prior art, and to jointly optimize the power and trajectory of UAVs in a NOMA (non-orthogonal multiple access) communication network to achieve the maximum energy efficiency of UAV communication coverage in the region.

[0006] To achieve the above technical objectives, the technical solutions adopted by the present invention are as follows:

[0007] A method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage, where the UAV periodically provides communication services to ground users within its jurisdiction. The method includes:

[0008] Step 1: Determine the initial periodic flight trajectory of the UAV based on the service time period of the UAV, the maximum speed constraint of the UAV, and the location information of each ground user as the initial fixed trajectory for Step 4.

[0009] Step 2: According to the free space path loss model of communication between the UAV and the users, obtain the channel gain between the UAV and each user at each moment. Sort the users in descending order of channel gain at each moment, and then calculate the transmission rate of the UAV.

[0010] Step 3: Use the Dinkelbach algorithm to convert the problem of maximizing the energy efficiency fraction of UAV communication coverage into a more convenient-to-solve and lower-complexity subtraction problem, obtain the corresponding objective function, and set the corresponding maximum speed constraint of the UAV, UAV energy consumption constraint, and minimum rate requirement constraint of the users.

[0011] Step 4: Based on Steps 2 and 3, set the optimized network structure of the user scheduling and user power allocation algorithm based on the dynamic programming algorithm. With the trajectory fixed, obtain more optimal power allocation and scheduling information, and update the network structure of regional communication.

[0012] Step 5: Based on Steps 2 and 3, set the optimized network structure of the trajectory optimization algorithm based on the successive convex approximation algorithm. With the power allocation and scheduling information fixed, obtain a more optimal trajectory, and update the network structure of regional communication.

[0013] Step 6: Iteratively update the network structure in Steps 4 and 5 to obtain the optimized network structure of regional communication, so as to maximize the energy efficiency of UAV communication coverage within the region under the conditions of meeting the maximum speed constraint of the UAV, UAV energy consumption constraint, and minimum rate requirement constraint of the users.

[0014] To optimize the above technical solution, the specific measures taken also include:

[0015] In the above Step 1, according to the period T of UAV service, the maximum speed constraint V of the UAV max , and the location information M = {M i , i = 1, 2,..., M} of each ground user, obtain the initial flight trajectory of the UAV, which specifically includes:

[0016] First, obtain the geometric center C of the ground user locations:

[0017]

[0018] After that, the distance d between the geometric center C and each user is calculated. i :

[0019] d i = ||C - M i ||

[0020] Through the distance d i The radius r of the smallest circle centered at C that can cover all users is obtained. u :

[0021] r u = max{d i , i = 1, 2,..., M}

[0022] Through the cycle T of the drone service and the maximum speed constraint V of the drone max The maximum allowable radius r is calculated. max :

[0023]

[0024] where π is the pi;

[0025] Through r u and r max The radius r of the initial flight path is obtained. tra :

[0026] r tra = min(r u , r max )

[0027] Through the geometric center C and the flight path radius r tra The initial flight path q of the drone is obtained. 0 [n]:

[0028] q 0 [n] = [x c + r tra cosθ n , y c + r tra sinθ n , n = 1,..., N.

[0029] q[n] is the position of the drone at the nth moment, x c is the abscissa of the geometric center C, y c is the ordinate of the geometric center C.

[0030] In the above step 2, the free space path loss model for the communication between the drone and the user is:

[0031]

[0032] Among them, g m,n represents the channel gain between the UAV and user m at time n;

[0033] ε represents the channel gain at d 0 = 1m;

[0034] d m,n represents the distance between the UAV and user m at time n;

[0035]

[0036] H is the flight altitude of the UAV.

[0037] In the above step 2, the downlink transmission rate calculation formula between the UAV and user m at time n is:

[0038]

[0039] Among them, γ m,n represents the downlink transmission rate between the UAV and user m at time n;

[0040] p m,n represents the power allocated to user m at time n;

[0041] σ 2 represents the additive white Gaussian noise power;

[0042] M n represents the set of users whose channel gain is better than that of user m among the users connected to the UAV at time n, and this user set includes user m.

[0043] The above step 3 specifically includes the following steps:

[0044] Step 301, the maximum energy efficiency fraction problem covered by UAV communication is modeled as the following maximization objective function:

[0045]

[0046]

[0047]

[0048] M n+1 = M 1

[0049]

[0050]

[0051]

[0052] Among them, U is the throughput of the UAV in the entire cycle;

[0053] E is the energy consumption model of the UAV;

[0054] y min is the minimum rate constraint of the user;

[0055] v max is the maximum speed constraint of the UAV;

[0056] a m,n is the user scheduling information, indicating that user m establishes a communication connection with the UAV at time n;

[0057] p max is the maximum power that the UAV can allocate;

[0058] Step 302, define it as U:

[0059]

[0060] Among them, B is the bandwidth of the UAV;

[0061] E(v[n]) is the energy consumption model of the UAV, defined as E(v[n]):

[0062]

[0063] Among them, P 0 is the blade profile power of the UAV in the hover state;

[0064] P i is the induced power of the UAV in the hover state;

[0065] U tip is the tip speed of the UAV rotor blade;

[0066] v 0 is the average rotor induced speed during hover;

[0067] d 0 is the fuselage drag ratio, and S is the rotor hardness;

[0068] ρ is the air density, and A is the rotor area;

[0069] v[n] is the flight speed of the UAV at time n;

[0070] Obtain the total energy consumption of the UAV in the cycle, defined as E:

[0071]

[0072] Step 303: Define \(t\) as the number of iterations, where \(t\) max represents the maximum number of iterations, and \(t\in\{1,2,\ldots,t\) max \(\}\), and we get:

[0073]

[0074] Step 304: By using the Dinkelbach algorithm, convert the fraction in the objective function into subtraction, and the objective function is re-expressed as:

[0075] \(\max U-\eta\) t-1 \(E\).

[0076] The user scheduling and user power allocation algorithm based on the dynamic programming algorithm described in Step 4 above is specifically as follows:

[0077] When the trajectory information is fixed, the energy consumption \(E\) of the UAV is also fixed. According to the objective function, it can be known that obtaining the maximum energy efficiency fraction of the UAV communication coverage is equivalent to obtaining the maximum throughput \(U\) of the UAV communication coverage. The way to obtain the maximum throughput \(U\) of the UAV communication coverage is as follows:

[0078] User sorting: At each moment, re-sort the users from high to low according to the channel gain;

[0079] Power allocation: Based on the idea of dynamic programming, at each moment, traverse and allocate power from high to low according to the channel gain to obtain the optimal throughput at each moment, and thus obtain the total optimal throughput.

[0080] In Step 5 above, when the power allocation and scheduling information are fixed, the objective function is a function only related to the trajectory. The trajectory optimization algorithm based on the successive convex approximation algorithm converts the objective function problem into a convex optimization problem and uses the convex optimization toolbox to solve it based on the successive convex approximation method, specifically including:

[0081] Step 501: Rewrite the UAV energy consumption model. When \(v[n]\gg v\) 0 \(\), through the first-order Taylor approximation formula the approximate expression of the energy consumption model can be obtained:

[0082]

[0083] At this point, the objective function is re-expressed as:

[0084] \(\max U-\eta\) t-1 \(E'\)

[0085] where \(\eta\) t-1 \(E'\) has become the form of a convex function;

[0086] Step 502: For U, rewrite the transmission rate formula:

[0087]

[0088] Step 503: The of the transmission rate formula m,n is a convex function with respect to the distance d. Obtain its first-order Taylor expansion expression at each flight path point of the UAV to get its lower bound, defined as:

[0089]

[0090] Step 504: For the of the transmission rate formula, introduce a new slack variable l, defined as:

[0091]

[0092]

[0093] Perform a first-order Taylor expansion expression for the slack variable constraint to obtain:

[0094]

[0095] Rewrite the second term of the transmission rate formula as:

[0096]

[0097] Step 505: Rewrite the minimum rate constraint as:

[0098]

[0099] After that, the entire objective function problem is transformed into a solvable convex optimization problem, and the successive convex approximation toolbox can be used to solve the problem to obtain a better flight path when the power allocation and scheduling information are fixed.

[0100] In the above step 6, consider a UAV that periodically provides communication services for users in the area. Denote the network structure of area communication as G=(M, Q, A, P), where M is the set of user location vectors, Q is the set of UAV flight paths, A is the set of user scheduling, and P is the set of user power allocation;

[0101] The current structure at each step is expressed as: G(A i ,P i ,Q i ) = U(A i ,P i ,Q i ) - η i-1E(Q i );

[0102] Among them, i is the iteration number of the i-th time. The optimized regional communication network structure is obtained through the following steps, specifically as follows:

[0103] Step 6.1: Optimize the network structure G(A i ,P i ,Q i ) = U(A i ,P i ,Q i ) - η i-1 E(Q i ), to obtain a better power allocation P i+1 and a better user scheduling A i+1 , and update the structure to: G(A i+1 ,P i+1 ,Q i ) = U(A i+1 ,P i+1 ,Q i ) - η i-1 E(Q i );

[0104] Step 6.2: Optimize the network structure G(A i +1 ,P i+1 ,Q i ) = U(A i+1 ,P i+1 ,Q i ) - η i-1 E(Q i ), to obtain a better UAV trajectory Q i+1 , and update the structure to: G(A i+1 ,P i +1 ,Q i+1 ) = U(A i+1 ,P i+1 ,Q i+1 ) - η i-1 E(Q i+1 );

[0105] Step 6.3: Obtain a better structure G(A i+1 ,P i+1 ,Q i+1 ) = U(A i+1 ,P i+1 ,Q i+1 ) - η i-1 E(Qi+1 ), and then update the value of η as follows: Optimize G(A i+1 , P i +1 , Q i+1 ) = U(A i+1 , P i+1 , Q i+1 ) - η i E(Q i+1 ) until G(A, P, Q) < ε, where ε is set to 0, and output the final regional communication network structure.

[0106] The present invention has the following beneficial effects: According to the free space path loss model of communication between the UAV and users, the channel gain between each user and the UAV at each moment is obtained. After sorting the users at each moment according to the channel gain, the user scheduling and power allocation of the UAV under a fixed flight path are obtained through dynamic programming; based on the user scheduling and power allocation of the UAV under a fixed flight path obtained by dynamic programming, the flight path problem is converted into a solvable convex optimization problem through the Dinkelbach algorithm, and the flight path under the fixed user scheduling and power allocation is obtained; according to the iterative optimization algorithm based on block coordinate descent, the user scheduling and power allocation of the UAV under a fixed flight path and the flight path under the fixed user scheduling and power allocation are iterated repeatedly to obtain the optimal user scheduling, optimal power allocation within the period, and the optimal flight path of the UAV within the period, so as to maximize the energy efficiency of the UAV's communication coverage in the area under the condition of satisfying the constraints between the UAV and users. BRIEF DESCRIPTION OF THE DRAWINGS

[0107] Figure 1 It is a schematic diagram of the UAV communication coverage system scenario involved in the present invention;

[0108] Figure 2 It is an experimental simulation result diagram of the number of users and energy efficiency under different methods;

[0109] Figure 3 It is an experimental simulation result diagram of the number of users served by the UAV in the same time slot of the present invention when the minimum user rate constraint is 3 * 10 7 bits / s;

[0110] Figure 4 It is an experimental simulation result diagram of the number of users served by the UAV in the same time slot of the present invention when the minimum user rate constraint is 6 * 10 7 bits / s;

[0111] Figure 5 It is an experimental simulation result diagram of the number of users served by the UAV in the same time slot of the present invention when the minimum user rate constraint is 10 * 10 7Experimental simulation result diagram of the number of users served by the drone in the same time slot under bits / s;

[0112] Figure 6 Experimental simulation result diagram of the number of iterations and energy efficiency of the present invention;

[0113] Figure 7 Experimental simulation result diagram of the optimized drone flight path planning of the present invention under different cycles;

[0114] Figure 8 Relationship diagram between different cycles and energy efficiency;

[0115] Figure 9 Relationship diagram between different cycles and drone energy consumption;

[0116] Figure 10 Relationship diagram between the maximum transmit power and energy efficiency under different methods;

[0117] Figure 11 Flowchart of the method for jointly optimizing user scheduling, power allocation, and drone flight path for NOMA communication coverage of the present invention. Detailed implementation manners

[0118] The following further describes the embodiments of the present invention in detail with reference to the accompanying drawings.

[0119] A method for jointly optimizing user scheduling, power allocation, and drone flight path for NOMA communication coverage realizes the maximization of energy efficiency of drone flight path planning by jointly optimizing power allocation and flight path optimization in the downlink wireless communication network of non - orthogonal multiple access (NOMA). The drone periodically provides communication services for ground users within its jurisdiction. According to the free - space path loss model of communication between the drone and users, the channel gain between each user and the drone at each moment is obtained. After sorting the users at each moment according to the channel gain, the user scheduling and power allocation of the drone under a fixed flight path are obtained through dynamic programming; based on the user scheduling and power allocation of the drone under a fixed flight path obtained by dynamic programming, the flight path problem is converted into a solvable convex optimization problem through the Dinkelbach algorithm and the flight path under the fixed user scheduling and power allocation is obtained; according to the iterative optimization algorithm based on block coordinate descent, the user scheduling and power allocation of the drone under a fixed flight path and the flight path under the fixed user scheduling and power allocation are iterated repeatedly to obtain the optimal user scheduling, optimal power allocation, and the optimal flight path of the drone within the cycle, so as to maximize the energy efficiency of the communication coverage of the drone in the area under the condition of satisfying the constraints between the drone and users.

[0120] In a specific embodiment, a drone periodically provides communication services to users within a region. Denote the network structure of the regional communication as G = (M, Q, A, P), where M is the set of user location vectors, Q is the set of drone flight trajectories, A is the user scheduling set, and P is the user power allocation set. By jointly optimizing the drone power allocation, user scheduling, and flight trajectory, the system energy efficiency of the wireless communication system is maximized under the constraints of energy and drone dynamics.

[0121] In the scenario of applying to drone-assisted communication, there are optimization problems in resource allocation and trajectory planning. Specifically, there are the following challenges: 1) In order to provide better quality of service to ground users, it is particularly important to plan the placement location of the drone or the flight trajectory information of the drone. A good flight trajectory plan can greatly improve the capacity of the entire system; 2) Due to the limitation of the on-board resources of the drone, considering the energy efficiency of the drone, that is, the system capacity under unit energy consumption, is particularly important; 3) Considering the characteristics of the NOMA wireless communication network, there needs to be a large difference in the power allocated to users. Therefore, power needs to be allocated to each user under the limitation of not exceeding the maximum power.

[0122] Specifically, as Figure 8 shown, the method for jointly optimizing user scheduling, power allocation, and drone trajectory for NOMA communication coverage of the present invention includes the following steps:

[0123] Step 1, determine the initial periodic flight trajectory of the drone as the initial fixed trajectory in Step 4 according to the time period of drone service, the maximum speed constraint of the drone, and the location information of each ground user;

[0124] In Step 1, according to the service period T of the drone, the maximum speed constraint V max of the drone, and the location information M = {M i , i = 1, 2,..., M} of each ground user, first obtain the geometric center C of the ground user locations:

[0125]

[0126] After that, calculate the distance d i between the geometric center C and each user:

[0127] d i = ||C - M i ||

[0128] Through the distance d i , obtain the radius r u of the smallest circle centered at C that can cover all users:

[0129] r u = max{di , where \(i = 1, 2, \ldots, M\)

[0130] Based on the cycle \(T\) of the drone service and the maximum speed constraint \(V\) of the drone max Calculate the maximum allowable radius \(r\) max :

[0131]

[0132] where \(\pi\) is the ratio of a circle's circumference to its diameter;

[0133] Based on \(r\) u and \(r\) max Obtain the radius \(r\) of the initial flight path tra :

[0134] \(r\) tra = min(\(r\) u , \(r\) max )

[0135] Based on the geometric center \(C\) and the flight path radius \(r\) tra Obtain the initial flight path \(q\) of the drone 0 [n]:

[0136] \(q\) 0 [n] = [x c + \(r\) tra cos\(\theta\) n , y c + \(r\) tra sin\(\theta\) n , \(n = 1, \ldots, N\).

[0137] \(q[n]\) is the position of the drone at time \(n\), \(x\) c is the abscissa of the geometric center \(C\), and \(y\) c is the ordinate of the geometric center \(C\).

[0138] Step 2: According to the free space path loss model of the communication between the drone and the user, obtain the channel gain between the drone and each user at each moment. Sort the users in descending order of the channel gain at each moment, and then calculate the transmission rate of the drone;

[0139] 1) The distance between the drone and user \(m\) at time \(n\) is defined as \(d\) m,n :

[0140]

[0141] \(H\) is the flight altitude of the drone;

[0142] According to the free space path loss model of the communication between the drone and the user, obtain the channel gain between the drone and user \(m\) at time \(n\), defined as \(g\) m,n:

[0143]

[0144] ε is the channel gain at d = 1m;

[0145] 2) In the NOMA wireless network, the successive interference cancellation (SIC) technique is adopted to eliminate the interference generated by other users sharing the subchannel at the same time, and the channel response with noise normalization (CRNN) is used to determine the decoding order. Therefore, in the NOMA wireless network, users with good channel conditions can eliminate the interference generated by users with poor channel conditions, that is, users are only affected by the interference generated by users whose channel gains are better than their own at the same time when being served. So the signal-to-noise ratio of user m at time n can be obtained, defined as SINR m,n :

[0146]

[0147] where M n represents the set of users with better channel gains than user m among the users connected to the UAV at time n, including user m, p m,n represents the power allocated to user m at time n, σ 2 represents the additive white Gaussian noise power;

[0148] After that, the downlink transmission rate of user m at time n is obtained through the Shannon formula, defined as γ m,n :

[0149] γ m,n = log 2 (1 + SINR m,n )

[0150] Step 3: Convert the maximum energy efficiency fraction problem of UAV communication coverage into a subtraction problem that is more convenient to solve and has lower complexity through the Dinkelbach algorithm, obtain the corresponding objective function, and set the corresponding UAV maximum speed constraint, UAV energy consumption constraint, and user minimum rate requirement constraint;

[0151] In Step 3, the maximum energy efficiency problem of UAV communication coverage is modeled as the following maximization model:

[0152]

[0153] s.t.

[0154]

[0155]

[0156] M n+1= M 1

[0157]

[0158]

[0159]

[0160] y min is the minimum rate constraint of the user, v max is the maximum speed constraint of the UAV, a m,n is the user scheduling information, indicating that user m establishes a communication connection with the UAV at time n, p max is the maximum available power that can be allocated to the UAV;

[0161] where U is the throughput of the UAV over the entire period, defined as U:

[0162]

[0163] B is the bandwidth of the UAV;

[0164] E(v[n]) is the energy consumption model of the UAV, defined as E(v[n]):

[0165]

[0166] P 0 is the blade profile power of the UAV in the hover state, P i is the induced power of the UAV in the hover state, U tip is the tip speed of the rotor blade of the UAV, v 0 is the average rotor induced speed during hover, d 0 is the fuselage drag ratio, S is the rotor stiffness, ρ is the air density, A is the rotor area, and v[n] is the flight speed of the UAV at time n;

[0167] The total energy consumption of the UAV over the period is obtained, defined as E:

[0168]

[0169] Define t as the number of iterations, t max represents the maximum number of iterations, t ∈ {1, 2, …, t max}, and we get:

[0170]

[0171] Through the Dinkelbach algorithm, the fraction in the objective function is transformed into subtraction to reduce its complexity. The objective function is re-expressed as:

[0172] max U - η t-1 E

[0173] Step 4: Based on Step 2 and Step 3, set the optimized network structure of the user scheduling and user power allocation algorithm based on the dynamic programming algorithm. When the flight path is fixed, obtain more optimal power allocation and scheduling information, and update the network structure of area communication;

[0174] In Step 4, according to the user power allocation algorithm based on the dynamic programming algorithm, when the flight path is fixed, more optimal power allocation and scheduling information is obtained. Its characteristic is that when the flight path information is fixed, the energy consumption E of the UAV is also fixed. Obtaining the maximum energy efficiency problem of UAV communication coverage can be equivalent to obtaining the maximum throughput U of UAV communication coverage. Considering the characteristics of the NOMA wireless communication network, that is, a user is only interfered by users whose channel gains are better than its own at the same moment. The specific steps are as follows:

[0175] User sorting: Re - sort the users from high to low according to the channel gain at each moment;

[0176] Power allocation: According to the characteristics of the NOMA wireless communication network, a user is only interfered by users with better channel gains than its own at the same moment. Based on the idea of dynamic programming, at each moment, traverse and allocate power from high to low according to the channel gain to obtain the optimal throughput at each moment, so as to obtain the total optimal throughput.

[0177] Step 5: Based on Step 2 and Step 3, set the optimized network structure of the flight path optimization algorithm based on the successive convex approximation algorithm. When the power allocation and scheduling information are fixed, obtain a more optimal flight path, and update the network structure of area communication; Through a series of variable substitutions, methods such as introducing slack variables, convert the original non - convex problem into a convex optimization problem, and use the convex optimization toolbox to solve it based on the successive convex approximation method.

[0178] In Step 5, when the power allocation and scheduling information are fixed, the objective function is a function only related to the flight path. First, rewrite the UAV energy consumption model. When v[n] >> v 0 , Through the first - order Taylor approximation formula The approximate expression of the energy consumption model can be obtained:

[0179]

[0180] So far, the objective function can be re - expressed as:

[0181] max U - η t-1 E′

[0182] where η t-1 E′ has been transformed into the form of a convex function. For U, rewrite the transmission rate formula as follows:

[0183]

[0184] For the formula is obviously a convex function with respect to the distance d m,n Perform a first-order Taylor expansion expression at each trajectory point of the UAV to obtain its lower bound, defined as:

[0185]

[0186] For the formula Introduce a new slack variable l, defined as:

[0187]

[0188]

[0189] Perform a first-order Taylor expansion expression for the slack variable constraint to obtain:

[0190]

[0191] Rewrite the second term of the formula as:

[0192]

[0193] Rewrite the minimum rate constraint as:

[0194]

[0195] After that, the entire problem is transformed into a solvable convex optimization problem, and the successive convex approximation toolbox can be used to solve the problem to obtain a better trajectory when the power allocation and scheduling information are fixed.

[0196] Step 6: Use a structure optimization algorithm based on the idea of the block coordinate descent algorithm to iteratively update the network structure in Steps 4 and 5, so as to maximize the energy efficiency of the UAV's communication coverage in the area under the conditions of satisfying the UAV's maximum speed constraint, UAV's energy consumption constraint, and user's minimum rate requirement constraint, and obtain the optimized regional communication network structure.

[0197] In Step 6, based on the optimization algorithms in Steps 4 and 5, use a structure optimization algorithm based on the idea of the block coordinate descent algorithm to iteratively optimize them, so as to maximize the energy efficiency of the UAV's communication coverage in the area under the conditions of satisfying the UAV's speed constraint, UAV's energy constraint, and user's minimum rate requirement constraint.

[0198] The network structure of area communication is \(G=(M, Q, A, P)\), and the current structure at each step is expressed as: \(G(A i ,P i ,Q i ) = T(A i ,P i ,Q i ) - \(\eta\ i-1 E(Q i ), where \(i\) is the number of the \(i\)-th iteration, and the optimized structure is obtained by the following steps, specifically as follows:

[0199] 6.1. Optimize the structure \(G(A i ,P i ,Q i ) = T(A i ,P i ,Q i ) - \(\eta\ i-1 E(Q i ) to obtain a better power allocation \(P i+1 and a better user scheduling \(A i+1 , and update the structure to: \(G(A i+1 ,P i+1 ,Q i ) = T(A i+1 ,P i+1 ,Q i ) - \(\eta\ i-1 E(Q i );

[0200] 6.2. Optimize the structure \(G(A i+1 ,P i+1 ,Q i ) = T(A i+1 ,P i+1 ,Q i ) - \(\eta\ i-1 E(Q i ) to obtain a better UAV trajectory planning \(Q i+1 , and update the structure to: \(G(A i+1 ,P i+1 ,Q i+1 ) = T(A i+1 ,P i+1 ,Q i+1 ) - \(\eta\ i-1 E(Q i+1 );

[0201] 6.3. Obtain a better structure \(G(A i+1 ,P i+1 ,Qi+1 ) = T(A i+1 , P i+1 , Q i+1 ) - η i-1 E(Q i+1 ), and then update the value of η as follows: Repeat the above steps for G(A i+1 , P i+1 , Q i+1 ) = T(A i +1 , P i+1 , Q i+1 ) - η i E(Q i+1 ) for optimization until G(A, P, Q) < ε, where ε is used to determine whether the algorithm converges, with a value of 0, and output the final regional communication network structure.

[0202] The method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage according to the present invention, Figure 1 is a scenario that maximizes the energy efficiency of UAV communication coverage in the region under the conditions of meeting UAV speed constraints, UAV energy constraints, and user minimum rate requirement constraints provided by the present invention.

[0203] Figure 2 is a graph showing the relationship between the number of users and energy efficiency under different methods. The methods from top to bottom are the DP - SCA method proposed in this patent, the trajectory optimization method under fixed power, the fixed trajectory method under power optimization, and the trajectory optimization method under fixed power in the time - division multiplexing case. It can be seen that in the case of any number of users, the energy efficiency of the method proposed in this patent for the entire system is the highest. As the number of users increases from 10 to 20, the energy efficiency growth increases from 17.3% to 19.9%. This is because as the number of users increases, more suitable users can be found to provide services during the iterative solution process to obtain greater energy efficiency. At the same time, as the number of users increases, the energy efficiency improves more and more, because non - orthogonal multiple access is more suitable for multi - user application scenarios. By comparing the trajectory optimization method under fixed power and the fixed trajectory method under power optimization, it can be seen that the trajectory has a great impact on the energy efficiency improvement of the entire system. A good trajectory can bring better overall energy efficiency to the system, and at the same time, power allocation and user scheduling are also essential for the system.

[0204] Figures 3 - 5 is a system structure diagram under different minimum user rates. The user minimum rate is from 3 * 10 7 bits / s, 6 * 10 7 bits / s, 10 * 10 7The bits / s increases sequentially, recording the system structures under different minimum user rates. In the figure, the triangles represent the coordinate positions of the ground users, the circles represent the coordinate positions of the UAVs at different time slots, and the dashed lines represent the communication links through which the UAVs provide services to the users at the current time slot. It can be seen that under the non-orthogonal multiple access technology, the UAV can provide downlink services to multiple users in one time slot. However, the number of users that the UAV serves simultaneously in one time slot decreases as the minimum user rate constraint increases. This is because the UAV has to meet the quality of service constraints of all users in each time slot, and the available bandwidth resources are limited. Therefore, the number of served users decreases.

[0205] Figure 6 This is the relationship diagram between the number of iterations and the energy efficiency of the method for jointly optimizing user scheduling, power allocation, and UAV trajectory in the NOMA communication coverage described in the present invention. It can be seen that as the number of iterations increases continuously, the energy efficiency of the entire system also increases. At the beginning of the algorithm iteration, the energy efficiency curve rises rapidly until the performance of the algorithm starts to converge. At the same time, the change in energy efficiency can reach convergence in 15 iterations, demonstrating that the method of the present invention has good convergence.

[0206] Figure 7 This is the optimized system structure diagram under different periods. The triangles represent the user positions, the square curves represent the optimized UAV flight trajectories under a period of 80 seconds, and the circular curves represent the optimized UAV flight trajectories under a period of 40 seconds. Figure 8 、 Figure 9 This gives the relationship between different periods and the system performance, where Figure 8 is the relationship diagram between different periods and the energy efficiency, Figure 9 is the relationship diagram between different periods and the UAV energy consumption. Figure 8 It can be seen from this that the DP-SCA iterative algorithm performs well at any period length, and the energy efficiency of the system increases as the period increases. This is because when the flight period is long, the UAV can better adjust its trajectory to approach and serve the ground users. In the case of only optimizing the power with a fixed trajectory, the energy efficiency of the system has been significantly improved after 50s. This is because when the period is too small, restricted by the maximum flight speed of the UAV, the UAV cannot approach all ground users to provide good services. At the same time, as Figure 9 shows, when the period is small, the UAV needs to fly at a higher speed to obtain a larger coverage area, so it will consume more energy. As the flight period increases, after all ground users can be covered, the flight speed of the UAV decreases and the energy consumption decreases. Therefore, the energy efficiency of the system will increase significantly. Figure 7It can be seen that when the time of a complete cycle is short, the trajectory of the UAV is flying towards the user. However, due to the limitation of the flight cycle, the distance is not very close. When the time of the complete cycle becomes longer, the UAV can fly a longer distance, so it will be closer to the user and better provide services to the user, ensuring the service quality of the user.

[0207] Figure 10 The relationship diagrams of the maximum transmission power and the energy efficiency under different methods are given. The methods from top to bottom are the DP-SCA iterative algorithm, the trajectory optimization algorithm under fixed power, and the fixed trajectory algorithm under power optimization. It can be seen that under the same transmission power, the DP-SCA iterative algorithm is higher than other schemes, and the system energy efficiency increases with the increase of the UAV transmission power. This is because as the transmission power increases, the connection quality between the UAV and the ground user becomes better, and the throughput of the system increases accordingly, so the system energy efficiency increases.

[0208] The above is only the preferred implementation mode of the present invention. The protection scope of the present invention is not limited to the above embodiments. All technical solutions within the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those of ordinary skill in the art of this technology, several improvements and refinements made without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage, where the UAV periodically provides communication services to ground users within its jurisdiction. Characterized in that, The method includes: Step 1, determine the initial periodic flight trajectory of the UAV as the initial fixed trajectory in Step 4 according to the time period of UAV service, the maximum speed constraint of the UAV, and the location information of each ground user. Step 2, according to the free space path loss model of communication between the UAV and the users, obtain the channel gain between the UAV and each user at each moment, sort the users in descending order of channel gain at each moment, and calculate the transmission rate of the UAV. Step 3, convert the maximum energy efficiency fraction problem of UAV communication coverage into a subtraction problem through the Dinkelbach algorithm, obtain the corresponding objective function, and set the UAV maximum speed constraint, UAV energy consumption constraint, and user minimum rate requirement constraint. Step 4, based on Step 2 and Step 3, set the optimization network structure of the user scheduling and user power allocation algorithm based on the dynamic programming algorithm. With the trajectory fixed, obtain better power allocation and scheduling information, and update the network structure of regional communication. Step 5, based on Step 2 and Step 3, set the optimization network structure of the trajectory optimization algorithm based on the successive convex approximation algorithm. With the power allocation and scheduling information fixed, obtain a better trajectory, and update the network structure of regional communication. Step 6, iteratively update the network structure in Step 4 and Step 5 to obtain an optimized regional communication network structure, so as to maximize the energy efficiency of UAV communication coverage within the region under the conditions of meeting the UAV maximum speed constraint, UAV energy consumption constraint, and user minimum rate requirement constraint.

2. A method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage according to claim 1. Characterized in that, In step 1, according to the cycle T of the drone service, the maximum speed constraint V of the drone max , and the location information M of each ground user i , i = 1, 2, …, M, the initial flight trajectory of the drone is obtained, specifically including: First, obtain the geometric center C of the ground user positions: After that, the distance d between the geometric center C and each user is calculated i : d i = ||C - M i || By distance d i Obtain the radius r of the smallest circle centered at C that can cover all users u : r u = max{d i , i = 1, 2, ..., M} Through the cycle T of the drone service and the maximum speed constraint V of the drone max The maximum allowable radius r is calculated max : where π is the pi. Through r u and r max obtain the radius r of the initial flight path tra : r tra = min(r u , r max ) Through the geometric center C and the flight path radius r tra Obtain the initial flight path q of the UAV 0 [n]: q 0 [n] = [x c + r tra cos θ n , y c + r tra sin θ n , n = 1, …, N. q[n] is the position of the UAV at time n, and x c is the abscissa of the geometric center C, and y c is the ordinate of the geometric center C.

3. A method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage according to claim 2. Characterized in that, In Step 2, the free space path loss model of communication between the UAV and the users is: where, g m,n represents the channel gain between the drone and user m at time n; ε represents the channel gain at d 0 = 1 m; d m,n represents the distance between the drone and user m at moment n; H is the flight altitude of the UAV.

4. A method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage according to claim 3. Characterized in that, In Step 2, the downlink transmission rate calculation formula between the UAV and user m at moment n is: where, γ m,n represents the downlink transmission rate between the UAV and user m at time n; p m,n represents the power allocated to user m at time n; σ 2 represents the power of additive white Gaussian noise; M n Denotes the set of users whose channel gain is better than that of user m among the users who establish a link with the UAV at time n. This user set includes user m.

5. A method for jointly optimizing user scheduling, power allocation, and UAV trajectory for NOMA communication coverage according to claim 4. Characterized in that, The specific steps of Step 3 are as follows: Step 301, the maximum energy efficiency fraction problem of UAV communication coverage is modeled as the following maximization objective function: M n+1 = M 1 where U is the throughput of the UAV in the entire period. E is the energy consumption model of the UAV. y min is the minimum user rate constraint; v max is the maximum speed constraint of the UAV; a m,n is user scheduling information, indicating that user m establishes a communication connection with the drone at time n; p max is the maximum power that can be allocated by the drone; Step 302, define U as: where B is the bandwidth of the UAV. E(v[n]) is the energy consumption model of the UAV, defined as E(v[n]): Among them, P 0 is the blade profile power of the UAV in the hovering state; P i is the induced power of the UAV in the hovering state; U tip is the tip speed of the drone rotor blade; v 0 is the average rotor induced velocity during hovering; d 0 where d is the fuselage drag ratio and S is the rotor hardness; ρ is the air density, A is the rotor area; v[n] is the flight speed of the drone at time n; The total energy consumption of the drone during the cycle is obtained, which is defined as E: Step 303, define t as the number of iterations, where t max represents the maximum number of iterations, and t ∈ {1, 2, …, t max}, and we get: Step 304: Using the Dinkelbach algorithm, the fraction in the objective function is converted into subtraction, and the objective function is re-expressed as: max U-η t-1 E。 6. A method for NOMA communication coverage that jointly optimizes user scheduling, power allocation, and drone trajectory according to claim 5, It is characterized in that The user scheduling and user power allocation algorithm based on the dynamic programming algorithm described in step 4 is specifically: When the track information is fixed, the energy consumption E of the UAV is also fixed. According to the objective function, the problem of obtaining the maximum energy efficiency score of the UAV communication coverage is equivalent to obtaining the maximum throughput U of the UAV communication coverage. The maximum throughput U of the UAV communication coverage is obtained as follows: User sorting: re-sort users from high to low according to channel gain at each moment; Power allocation: Based on the idea of ​​dynamic programming, at each moment, the power is allocated from high to low channel gain to obtain the optimal throughput at each moment, thereby obtaining the overall optimal throughput.

7. A method for NOMA communication coverage that jointly optimizes user scheduling, power allocation, and drone trajectory according to claim 6, It is characterized in that In step 5, when the power allocation and scheduling information are fixed, the objective function is a function related only to the track. The track optimization algorithm based on the successive convex approximation algorithm converts the objective function problem into a convex optimization problem, and solves it using the convex optimization toolbox based on the successive convex approximation method, specifically including: Step 501: rewrite the energy consumption model of the drone to obtain an approximate expression of the energy consumption model, which is defined as: At this point, the objective function is reformulated as: max U-η t-1 E ′ where η t-1 E ′ has become a convex function form; Step 502: For U, rewrite the transmission rate formula: Step 503, for the transmission rate formula is a convex function with respect to the distance d m,n Perform a first-order Taylor expansion expression at each trajectory point of the UAV to obtain its lower bound, which is defined as: Step 504. For the transmission rate formula introduce a new slack variable l, which is defined as: Performing a first-order Taylor expansion for the slack variable constraint yields: The second term of the transmission rate formula is rewritten as: Step 505: rewrite the minimum rate constraint as follows: The entire objective function problem is then transformed into a solvable convex optimization problem, and the successive convex approximation toolbox can be used to solve the problem to obtain a better trajectory when the power allocation and scheduling information are fixed.

8. A method for NOMA communication coverage that jointly optimizes user scheduling, power allocation, and drone trajectory according to claim 7, It is characterized in that In step 6, consider a drone that periodically provides communication services to users in the area. The network structure of regional communication is G = (M', Q, A, P), where M' is the user position vector set, Q is the drone track set, A is the user scheduling set, and P is the user power allocation set; The current structure at each step is represented as: G(A i , P i , Q i ) = U(A i , P i , Q i ) - η i-1 E(Q i ); Where i is the number of iterations for the i-th time. The optimized regional communication network structure is obtained by the following steps, as follows: Step 6.

1. Through the optimized network structure \(G(A i ,P i ,Q i ) = U(A i ,P i ,Q i ) - η i-1 E(Q i ), a better power allocation \(P i+1 and a better user scheduling \(A i+1 are obtained. The structure is updated to: \(G(A i+1 ,P i+1 ,Q i ) = U(A i+1 ,P i+1 ,Q i ) - η i-1 E(Q i ); Step 6.

2. Through the optimized network structure G(A i+1 ,P i +1 ,Q i ) = U(A i+1 ,P i+1 ,Q i ) - η i-1 E(Q i ), a better UAV trajectory Q i+1 is obtained, and the structure is updated to: G(A i+1 ,P i+1 ,Q i +1 ) = U(A i+1 ,P i+1 ,Q i+1 ) - η i-1 E(Q i+1 ); Step 6.3: Obtain a better structure G(A i+1 ,P i+1 ,Q i+1 ) = U(A i+1 ,P i+1 ,Q i+1 ) - η i-1 E(Q i+1 ). Then update the value of η as follows: Optimize G(A i+1 ,P i+1 ,Q i+1 ) = U(A i+1 ,P i+1 ,Q i+1 ) - η i E(Q i+1 ) according to the above update steps until G(A, P, Q) < ε, where ε takes the value of 0, and output the final regional communication network structure.

Images