An optimization method for cognitive UAV trajectory and resource allocation

Abstract

The disclosed embodiments are about a method for optimizing the trajectory and resource allocation of a cognitive UAV. The method includes: constructing a probabilistic line-of-sight link model based on a cognitive UAV communication network; establishing a compromise optimization model according to the probabilistic line-of-sight link model; decomposing the compromise optimization model into a two-layer problem, solving the two-layer problem using a first algorithm and a second algorithm, and obtaining the global optimal throughput. On the one hand, the disclosed embodiments explore the optimal dynamic trajectory, access strategy, and power allocation of UAVs in a probabilistic line-of-sight channel under the severe co-channel interference constraints of the primary network, thereby improving the performance of the secondary network. On the other hand, the initial problem is decomposed into a two-layer problem, and an improved particle swarm algorithm is proposed to effectively achieve a compromise between maximizing throughput and minimizing co-channel interference.

Description

Technical Field

[0001] The disclosed embodiments relate to the field of wireless communication technology, and more particularly to a method for optimizing cognitive drone trajectories and resource allocation. Background Art

[0002] Unmanned aerial vehicle (UAV) communications can significantly improve the performance and coverage of ground communication networks. Due to its flexibility, it can be widely deployed in many scenarios, such as post-disaster recovery and emergency rescue. In addition, UAVs can also increase the reliability of traditional communication systems, such as acting as base stations or relays for users beyond the coverage of ground networks, providing more reliable wireless services.

[0003] Under current conditions, UAVs usually do not have pre-allocated spectrum, but share spectrum resources with other ground wireless devices. However, the current static spectrum allocation policy makes the spectrum shortage problem more acute. Therefore, cognitive radio (CR) technology is regarded as one of the important technologies to alleviate spectrum problems through dynamic spectrum sharing. In the CR network, UAVs can dynamically access the authorized spectrum of the main network as secondary users. Obviously, compared with the traditional ground CR network, UAVs act as perception and communication nodes as secondary users, and have stronger line of sight (LoS) links and deployment flexibility compared to traditional ground CR networks. In the existing technology, only simple LoS links are considered, which is obviously not rigorous in urban communication scenarios, especially in urban environments with high building density.

[0004] Therefore, it is necessary to improve one or more problems existing in the above-mentioned related technical solutions.

[0005] It should be noted that this section is intended to provide background or context for the technical solutions of the present disclosure stated in the claims. The description herein is not admitted to be prior art by virtue of being included in this section. Summary of the invention

[0006] The purpose of the embodiments of the present disclosure is to provide a method for optimizing cognitive drone trajectories and resource allocation, thereby overcoming one or more problems caused by limitations and defects of related technologies, at least to a certain extent.

[0007] According to an embodiment of the present disclosure, a method for optimizing the trajectory and resource allocation of a cognitive drone is provided, comprising:

[0008] Constructing a probabilistic line-of-sight link model based on cognitive UAV communication network;

[0009] Establishing a compromise optimization model according to the probabilistic line-of-sight link model;

[0010] The compromise optimization model is decomposed into a two-layer problem, and the two-layer problem is solved using a first algorithm and a second algorithm to obtain a global optimal throughput.

[0011] In one embodiment of the present disclosure, the process of decomposing the compromise optimization model into a two-layer problem and solving the two-layer problem using a first algorithm and a second algorithm includes:

[0012] Decomposing the compromise optimization model into an access strategy problem between the UAV and the ground user and a joint optimization problem between the UAV trajectory and power allocation;

[0013] Solving the access strategy problem between the UAV and the ground user by using the first algorithm;

[0014] The second algorithm is then used to solve the joint optimization problem of the UAV trajectory and power allocation.

[0015] In one embodiment of the present disclosure, the first algorithm is an enumeration method, and the second algorithm is an improved particle swarm algorithm.

[0016] In one embodiment of the present disclosure, the improved particle swarm algorithm includes:

[0017] The inertia weight is used as a variable updated with the number of iterations to adjust the local and global search capabilities of particles;

[0018] The position of the particle swarm is initialized, and then the single particle coordinate quadratic decomposition method is used to improve the quality of the solution and the convergence speed. Only the position of the two-dimensional variables of each particle is updated each time to solve the problem that particles are easy to jump out of the feasible solution space in the later stage of the algorithm.

[0019] In one embodiment of the present disclosure, the drone communication network includes:

[0020] primary and secondary networks;

[0021] The primary network includes a primary user transmitter and a plurality of primary user receivers, and the primary user receivers are distributed in a circular area with the primary user transmitter as the center;

[0022] The secondary network includes a secondary user transmitter and a plurality of secondary user receivers, wherein the secondary user transmitter is a central UAV and the secondary user receivers are ground users;

[0023] Among them, the position of the main receiver is random, and the secondary network adopts the access mode of the infrastructure layer, that is, the drone guarantees the service quality of the main receiver by controlling the transmission power, thereby sharing the spectrum.

[0024] In one embodiment of the present disclosure, the drone acts as a monitor and flies along a planned trajectory, and transmits monitoring data back to the ground user at specific time intervals during the flight.

[0025] In one embodiment of the present disclosure, the drones are all equipped with a global positioning system to monitor their positions and dynamics in real time.

[0026] In one embodiment of the present disclosure, the trade-off optimization model includes:

[0027] Under the constraints of preset mission time, flight speed and primary user interference threshold, the expected average throughput of the UAV is maximized by optimizing the dynamic trajectory, power allocation and access strategy of the UAV.

[0028] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects:

[0029] In the embodiments of the present disclosure, through the above-mentioned optimization method of cognitive drone trajectory and resource allocation, on the one hand, under the severe co-channel interference constraint of the primary network, the optimal dynamic trajectory, access strategy and power allocation of the drone in the probabilistic line-of-sight channel are explored, and the performance of the secondary network is improved. On the other hand, the initial problem is decomposed into a two-layer problem, and an improved particle swarm algorithm is proposed to effectively achieve a compromise between maximizing throughput and minimizing co-channel interference. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0031] Figure 1 A step diagram of the optimization method for recognizing drone trajectories and resource allocation disclosed in the present invention is shown;

[0032] Figure 2 A three-dimensional system model diagram of a cognitive drone communication network in an exemplary embodiment of the present disclosure is shown;

[0033] FIG3( a ) shows a diagram of the optimal trajectory result of the UAV under a given time constraint using the improved particle swarm algorithm in an exemplary embodiment of the present disclosure;

[0034] FIG3( b ) shows a trajectory result diagram obtained by independent optimization in an exemplary embodiment of the present disclosure;

[0035] FIG4( a ) shows a diagram of the allocation strategy and transmission power results obtained by optimizing the improved particle swarm algorithm in an exemplary embodiment of the present disclosure;

[0036] FIG4( b ) shows a diagram of the final optimized allocation strategy and transmission power result obtained by independent optimization in an exemplary embodiment of the present disclosure;

[0037] Figure 5 A throughput comparison chart of the optimization algorithm of the present disclosure and other existing algorithms in an exemplary embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0038] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the disclosure will be more comprehensive and complete and to fully convey the concepts of the example embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0039] In addition, the accompanying drawings are only schematic illustrations of the embodiments of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated descriptions will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0040] In this example implementation, a method for optimizing the trajectory and resource allocation of a cognitive drone is first provided. Figure 1 As shown in , the optimization method of cognitive drone trajectory and resource allocation may include: steps S101 to S103.

[0041] Step S101: constructing a probabilistic line-of-sight link model based on the cognitive UAV communication network;

[0042] Step S102: establishing a compromise optimization model according to the probabilistic line-of-sight link model;

[0043] Step S103: decomposing the compromise optimization model into a two-layer problem, solving the two-layer problem using the first algorithm and the second algorithm, and obtaining a global optimal throughput.

[0044] Through the above-mentioned optimization method of cognitive drone trajectory and resource allocation, on the one hand, under the constraint of co-channel interference of the primary network, the optimal dynamic trajectory, access strategy and power allocation of drones in probabilistic line-of-sight channels are explored, and the performance of the secondary network is improved. On the other hand, the initial problem is decomposed into a two-layer problem, and an improved particle swarm algorithm is proposed to effectively achieve a compromise between maximizing throughput and minimizing co-channel interference.

[0045] Next, we will refer to Figure 1 to Figure 2Each step of the above-mentioned cognitive drone trajectory and resource allocation optimization method in this example implementation is described in more detail.

[0046] Step S101: Construct a probabilistic line-of-sight link model based on the cognitive UAV communication network.

[0047] Specifically, the cognitive drone communication network based on urban scenarios considered in this disclosure is as follows: Figure 2 As shown in the figure, the UAV acts as a monitor and flies along the planned trajectory. During the flight, it transmits the monitoring data back to the ground user at specific time intervals. The main network consists of a primary transmitter (PT) and K primary receivers (PR), with P and R k (k∈K={1,…,K}); the secondary network consists of a single UAV as a secondary transmitter and M ground users as secondary receivers, represented by S and G respectively. m (m∈M={1,…,M}). The position of PR is random, and the secondary network adopts the Underlay access mode, that is, the UAV controls the transmission power to ensure the quality of service (QoS) of the primary receiver, thereby achieving the purpose of spectrum sharing. The goal of this spectrum sharing scheme is to transmit observation data as much as possible while minimizing co-channel interference, thus sacrificing fairness for ground users. Each UAV is equipped with a global positioning system to monitor its position and dynamics in real time.

[0048] The location of the ground wireless user is given by G = {g 1 ,…,g M} T Indicates that The UAV needs to be within a given mission time T tot The ground monitoring and reporting from the starting point to the end point are completed within 10 minutes, and the flight altitude is h. The total time is T tot It is divided into N time slots, each time slot is T s The position of the UAV projected onto the horizontal plane at the nth time slot is It is represented by, where n∈N={1,…,N}. The distance from the UAV to the mth ground user is This paper comprehensively considers the ground-to-ground (G2G) channel and A2G channel. In order to be more in line with the actual scenario, random shadow fading is modeled as a probabilistic LoS link model. The channel gain between nodes u and v can be expressed as

[0049]

[0050]

[0051] in and They represent the path loss of the LoS link and the non-line-of-sight link (non-LoS, NLoS), respectively. uv is the distance between nodes u and v. represents the normalized small-scale fading, and In this paper, small-scale fading is ignored for simplicity because the trajectory of the UAV is mainly designed in an offline manner and its purpose is not to adapt to the randomness of small-scale fading. The average path loss of the LoS link and the NLOS link can be expressed as

[0052]

[0053]

[0054] where ξ LoS and NLoS denote the average additional loss of LoS link and NLoS link respectively, f is the carrier frequency, and c is the speed of light. The probability that the mth ground user and UAV are in LoS channel in time slot n is It usually depends on the propagation environment and the elevation angle between the node and the UAV, which can be expressed as

[0055]

[0056] Where α and β are constants determined by the propagation environment, and the elevation angle (in degrees) can be expressed as

[0057]

[0058] Let λ m [n]∈{0,1} represents the access strategy between UAV and ground users in time slot n, where λ m [n] = 1 means that the mth ground user accesses and communicates with the UAV, otherwise λ m [n] = 0. Since the UAV communicates with ground users in time-division multiple access (TDMA) mode, there is

[0059]

[0060] The transmission power of PT is P P It means that the received power at PR is Based on the uncertainty of PR position, this paper defines the protection boundary L * To ensure the QoS of the primary network. *It is determined by the PR signal-to-noise ratio (SNR) threshold, which is expressed as (σ 2 is the noise power at the PR). The location of the PR is usually unknown and has mobility and uncertainty. Therefore, this paper assumes that the PR is deployed at the protection boundary to maximize its QoS. In addition, the primary user interruption threshold is defined as follows

[0061]

[0062] Where P[n] is the transmit power of the UAV, is the LoS link probability between the mth UAV and PR in time slot n, θ out is the interrupt threshold.

[0063] Step S102: establishing a compromise optimization model according to the probabilistic line-of-sight link model.

[0064] Specifically, based on the probability of the existence of the primary user, we first calculate the expected throughput of the n-time slot system. It is expressed as follows

[0065]

[0066] Where k∈K={0,1}, k=1 means PT exists, otherwise k=0,π 1 and π 0 Based on the probabilistic LoS link constructed in the previous section, the expected throughput between the UAV and the mth ground user in time slot n is for

[0067]

[0068] in and are the channel capacities of the UAV and the mth ground user under the LoS link and the NLoS link, respectively, which can be expressed as

[0069]

[0070]

[0071] Since the UAV position and the probability of LoS link change over time, we consider the expected average in all time slots, which can be defined as

[0072]

[0073] The goal of this paper is to maximize the average expected throughput under constraints such as completion time, maximum speed, and interruption threshold. Therefore, the original problem can be modeled as

[0074]

[0075] stT≤T tot (15)

[0076] ‖Q[n+1]-Q[n]‖≤v max ·T s (16)

[0077]

[0078] (7),(8) (18)

[0079] Where T tot is the task completion time constraint, Q[0] = Q I ,Q[n+1]=Q F ,Q I , They represent the initial position and final position of the UAV projected onto the horizontal plane.

[0080] Step S103: decomposing the compromise optimization model into a two-layer problem, solving the two-layer problem using the first algorithm and the second algorithm, and obtaining a global optimal throughput.

[0081] Problem (P1) is a mixed integer non-convex optimization problem, in which the constraints defined by equations (7) and (16) are non-convex, so it is difficult to solve. In order to make the problem easier to solve, it is first transformed into a two-level optimization problem, that is, first solve the access strategy problem of UAV and ground users. The first-level problem is expressed as

[0082]

[0083] st(17),(18) (20)

[0084] It can be seen that the first level problem (P2.1) is a single variable integer problem, which can be solved by enumeration method. Indicates that, where λ m The value of (n) changes with the position of the UAV. Subproblems (P2.1) and (P2.2) make the original optimization problem tractable again after being decomposed. Then, the joint optimization problem of UAV trajectory and power allocation can be defined as

[0085]

[0086] st(15),(16) (22)

[0087] To solve the above problems, an improved algorithm based on the classical particle swarm algorithm is proposed to preset the initial particle coordinates, so as to obtain the optimal trajectory.

[0088] The variable search space in problem (P2.2) is 2N-dimensional, and the generator matrix is ​​given by Given, where A is the population of particles and t is the number of iterations. Therefore, corresponds to the position of the ath particle in the tth iteration. According to constraint (16), the particle swarm position is initialized as

[0089]

[0090] in represents the maximum integer not greater than (·), the fitness function in each iteration The average throughput expected and the penalty function based on constraint (16) Determine that the expressions of F(x) and g(x) are as follows

[0091]

[0092]

[0093] The value of ξ needs to be set to a large negative number. For each particle, its optimal position and fitness value are updated in each iteration as and Finally, the global optimal particle position q is obtained * , and the global optimal throughput y is obtained accordingly * . For a specific optimization problem (P2.2), this paper also makes the following improvements to the particle swarm algorithm to improve the algorithm's search ability. First, the proposed optimization algorithm uses the inertia weight w as a variable updated with the number of iterations to adjust the local and global search capabilities of the particles. Secondly, for the problem that particles are prone to jump out of the feasible solution space in the later stage of the algorithm, we use the method of single particle coordinate quadratic decomposition to improve the quality of the solution and the convergence speed. The detailed steps of the improved algorithm are given in Table 1. Where Ψ and Φ are matrices with elements randomly distributed in the range [0,1], t max Indicates the maximum number of iterations.

[0094] Table 1: Improved particle swarm algorithm

[0095]

[0096] This embodiment is further described below in conjunction with a specific simulation example.

[0097] This section gives the simulation results to verify the performance of the proposed algorithm. The channel model used has been given in Section 2. The simulation parameters are set as follows: A = 1000, t max =2×10 4 , B=10 5 Hz, P P =30dB,σ 2 =-78dBm, G = [-50, -300; 50, 150; 100, 0], N = 15, f = 2.44GHz and v max =18m / s. According to the regulations of the Federal Aviation Administration of the United States, the flight altitude of the drone is set at 100m. The channel environment is set to an urban environment with high building density, where α = 25, β = 0.112, ξ LoS =2.3dB and ξ NLoS =34dB. In the cognitive radio access mode based on the Underlay mode, the parameter is set to π 1 =0.7,γ th =5.5dB and θ out =-28dB.

[0098] Figure 3(a) shows the optimal trajectory of the UAV under a given time constraint using the optimization algorithm disclosed in this disclosure, and Figure 3(b) shows the comparison of the trajectories of independent optimization. Obviously, when the UAV is vertically above the ground user, the best channel and throughput will be obtained. Therefore, under the four different time constraints, the UAV tends to fly towards the ground user. When the mission time is relatively tight (such as T tot =35s), the UAV will first meet the time constraint and complete the task at the expense of throughput. When a more relaxed task completion time is assigned (such as T tot =50s), the UAV will stay at the ground user 3 for a longer time to obtain the best channel state and communication duration under the condition of being far away from the PR. However, in Figure 3(b), due to independent optimization, the UAV will not fly away from the main network to reduce the co-channel interference. This shows the necessity of the proposed joint optimization algorithm, which can maximize the average value of the throughput expectation of the UAV under the co-channel interference and time constraints.

[0099] Figure 4(a) shows the access strategy and power allocation of UAV and ground users under the corresponding trajectory of the optimization algorithm disclosed in the present invention, and Figure 4(b) shows the result of independent optimization. Obviously, in order to ensure the QoS of PR, the transmission power of UAV needs to be strictly controlled throughout the flight. When the UAV flies towards the ground user 2, the power drops faster. This is why the UAV is more inclined to fly towards user 1 rather than user 2 when the time constraints are 45s and 50s. When the UAV approaches user 2, the reduced distance will reduce its transmission power while increasing the channel gain. The trajectory in Figure 4(a) reflects the trade-off between the increased channel gain and the reduced power allocation of the UAV, while for Figure 4(b), the UAV flies directly above the ground user 2 as expected, and the resulting power loss deteriorates its throughput performance.

[0100] final, Figure 5 The proposed joint optimization scheme is compared with four benchmark schemes with the same complexity. Since the simulation results show that the algorithms of several schemes can generally converge at the 6000th iteration, the iteration parameters are set to A=1000, t max =6000. Among them, the performance of the straight trajectory is the worst; the circular trajectory is not suitable for the communication system proposed in the present invention because its radius increases with the flight time; the hover-flight strategy can obtain performance similar to that of the proposed algorithm, but there is still a certain gap; the independent optimization has a slightly worse performance than the hover-flight strategy because it does not control the UAV power allocation in real time. In summary, the performance of the proposed optimization scheme is better than the existing scheme, and the improved algorithm is better than the traditional particle swarm algorithm, which is superior and innovative.

[0101] This paper studies the efficient spectrum sharing strategy in cognitive UAV communication networks, and the joint optimization problem of cognitive UAV trajectory and resource allocation based on time constraints under the uncertainty of PR position. Under the constraint of co-channel interference of the primary network, the optimal dynamic trajectory, access strategy and power allocation of UAV in probabilistic LoS channel are explored to improve the performance of secondary network. The initial non-convex problem is decomposed into a two-layer problem, and an improved PSO algorithm is proposed to effectively achieve the compromise between maximizing throughput and minimizing co-channel interference. Simulation and numerical analysis verify the superiority of the proposed scheme compared with the traditional scheme.

[0102] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0103] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine different embodiments or examples described in this specification.

[0104] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

Claims

1. An optimization method for cognitive drone trajectory and resource allocation, It is characterized in that include: Constructing a probabilistic line-of-sight link model based on cognitive UAV communication network; Establishing a compromise optimization model according to the probabilistic line-of-sight link model; Decomposing the compromise optimization model into a two-layer problem, solving the two-layer problem using a first algorithm and a second algorithm, and obtaining a global optimal throughput, including: Decomposing the compromise optimization model into an access strategy problem between the UAV and ground users and a joint optimization problem between the UAV trajectory and power allocation; Solve the access strategy problem between the drone and the ground user by using the first algorithm; Then, the joint optimization problem of the UAV trajectory and power allocation is solved by the second algorithm; The first algorithm is an enumeration method, and the second algorithm is an improved particle swarm algorithm; The improved particle swarm algorithm comprises: The inertia weight is used as a variable updated with the number of iterations to adjust the local and global search capabilities of particles; Initialize the position of the particle swarm, and then use the single particle coordinate quadratic decomposition method to improve the quality of the solution and the convergence speed. Only the position of the two-dimensional variables of each particle is updated each time to solve the problem that particles are easy to jump out of the feasible solution space in the later stage of the algorithm. Wherein, the compromise optimization model includes: Under the constraints of preset mission time, flight speed and primary user interference threshold, the expected average throughput of the UAV is maximized by optimizing the dynamic trajectory, power allocation and access strategy of the UAV.

2. According to the optimization method of cognitive drone trajectory and resource allocation according to claim 1, It is characterized in that The UAV communication network includes: primary and secondary networks; The primary network includes a primary user transmitter and a plurality of primary user receivers, and the primary user receivers are distributed in a circular area with the primary user transmitter as the center; The secondary network includes a secondary user transmitter and a plurality of secondary user receivers, wherein the secondary user transmitter is a central UAV and the secondary user receivers are ground users; Among them, the position of the primary user receiver is random, and the secondary network adopts the Underlay access mode, that is, the UAV controls the transmission power to ensure the service quality of the primary user receiver, thereby sharing the spectrum.

3. According to the optimization method of cognitive drone trajectory and resource allocation as described in claim 2, It is characterized in that The UAV flies along a planned trajectory as a monitor, and transmits monitoring data back to the ground user at specific time intervals during the flight.

4. According to the optimization method of cognitive drone trajectory and resource allocation as described in claim 3, It is characterized in that The drones are all equipped with global positioning systems to monitor their positions and dynamics in real time.

Images