A method for computing, communicating and controlling a networked unmanned aerial vehicle and a system thereof

Abstract

The application discloses a kind of network connection unmanned plane's calculation, communication and control method and its system, wherein network connection unmanned plane's calculation, communication and control method, specifically includes the following steps: initialization state information;Carry out queue processing;Unmanned plane-base station pairing is carried out;Communication and calculation resource allocation are carried out;Unmanned plane trajectory optimization is carried out;Whether to converge to preset accuracy or iteration number reaches maximum iteration number is judged;If converge to preset accuracy or iteration number reaches maximum iteration number, output current time slot result;Whether to optimize all time slots is judged;If optimization completes all time slots, then the process is ended.The application breaks through the drawbacks of traditional network connection unmanned plane mobile edge computing system offline optimization method, and proposes an online optimization method.A calculation, communication and control method for network connection unmanned plane, and a system implementation thereof are proposed.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of mobile communication, in particular, to a computing, communication and control method and system for cellular-connected unmanned aerial vehicle. BACKGROUND

[0002] With the rise of the Internet of Things, the number of different types of ground terminal devices (TDs) such as cloud sensors, smart phones, wearable devices is growing rapidly; smart applications such as face recognition, interactive games, virtual reality are also emerging. However, due to the weak computing power and low battery capacity of these terminal devices, it is a difficult problem to effectively compute the large amount of computing data generated by these applications. Under this background, mobile edge computing (MEC) as a technology with development potential can help TDs to perform data processing and computing at the edge of the network, thereby solving the above problems. With flexible mobility, unmanned aerial vehicles (UAVs) have received more and more research and attention in recent years. Integrating unmanned aerial vehicles into existing cellular networks to form cellular-connected UAVs can achieve high data transmission rate and high reliable and safe performance by using the licensed frequency band of cellular networks. Therefore, cellular-connected UAVs not only can expand the service range of unmanned aerial vehicles, but also can effectively control in time. This has given rise to many promising new applications, such as package delivery, power inspection, environmental monitoring, etc. Then, these new applications are also closely related to edge computing technology. Cellular-connected UAVs can provide large-scale accurate and efficient services, and the network needs to provide computing power support in addition to providing communication connection to ensure that the unmanned aerial vehicle can make accurate decisions and predictions based on the collected information, so as to achieve satisfactory service quality. Considering that the unmanned aerial vehicle will encounter various environmental influences and various uncertain factors of the network during the execution of the task, the computing task generated thereby has randomness, and the computing task generated at the future time cannot be predicted from the current time, which brings serious challenges to the design of the actual system.

[0003] Therefore, how to provide an online optimization method which can be used for computing, communication and control of cellular-connected UAVs and system implementation is an urgent problem for those skilled in the art. SUMMARY

[0004] The application provides a networked unmanned aerial vehicle computing, communication and control method, specifically comprising the following steps: initializing state information; in response to the initialized state information, performing queue processing; in response to the completion of the queue processing, performing unmanned aerial vehicle-base station pairing; in response to the completion of the unmanned aerial vehicle-base station pairing, performing communication and computing resource allocation; in response to the completion of the communication and computing resource allocation, performing unmanned aerial vehicle trajectory optimization; in response to the completion of the unmanned aerial vehicle trajectory optimization, judging whether the preset precision is converged or the iteration number reaches the maximum iteration number; if the preset precision is converged or the iteration number reaches the maximum iteration number, outputting a current time slot result; in response to the output of the current time slot result, judging whether all time slots are optimized; if all time slots are optimized, the process ends.

[0005] The above, wherein the initialized state information comprises initializing the unmanned aerial vehicle flight trajectory, the unmanned aerial vehicle maximum transmission power P k , the total number of time slots, the unmanned aerial vehicle maximum CPU computing frequency, the system bandwidth B, the time slot length δ t , and the convergence precision and the maximum iteration number.

[0006] The above, wherein the queue processing comprises processing the computing queues at the unmanned aerial vehicles and the base stations respectively.

[0007] The above, wherein the change of the queue length at the unmanned aerial vehicle k Q k [n+1] is represented as:

[0008]

[0009] wherein Q k [n] represents the queue length of the unmanned aerial vehicle k at the time slot n, M represents the number of ground base stations, A k [n] represents the computing task amount generated by the unmanned aerial vehicle k at the time slot n, represents the computing task amount of the unmanned aerial vehicle k at the time slot n, f k [n] represents the allocated computing resource, c represents the number of computing periods required for computing 1 bit of data, δ t represents the length of one time slot, represents the receiving task amount of the base station m within one time slot.

[0010] The above, wherein the receiving task amount of the base station m within one time slot is represented as:

[0011]

[0012] wherein δ t represents the length of one time slot, γ k,m [n]=β0(||u k [n]-w m|| 2 +H 2 ) -α Hkm[n] is the channel coefficient between UAV k and base station m in time slot n, β0is the reference channel power gain, α is the path loss exponent, u k [n] represents the trajectory of UAV k in time slot n, H represents the flight height of the UAV, w m km[n] represents the location of base station m, ||.|| represents the Euclidean norm.

[0013] As above, wherein the change of the queue length of base station m Z k,m [n+1] is represented as:

[0014]

[0015] wherein Z k,m [n] represents the queue length of base station m in time slot n, D k,m [n] = δ t f k,m [n] / c represents the amount of tasks calculated by base station m for UAV k, δ t represents the length of a time slot, f k,m [n] represents the computing resource allocated by base station m, and c represents the number of computing periods required for each 1 bit of data.

[0016] As above, wherein, in the process of UAV-base station pairing, the optimal pairing scheme is obtained by the following expression:

[0017]

[0018] As above, wherein, wherein wherein Q k [n] represents the queue length of UAV k in time slot n, Z k,m [n] represents the queue length of base station m in time slot n, δ t represents the length of a time slot, B k,m [n] represents the communication bandwidth allocated by the system in time slot n for UAV k to offload tasks to base station m, p k [n] represents the transmission power of the UAV in each time slot, γ k,m [n] is the channel coefficient between UAV k and base station m in time slot n, and N0represents the white noise power spectral density at the base station end.

[0019] As above, wherein the communication and computing resource allocation includes obtaining the optimal UAV computing resource allocation, the optimal UAV computing resource allocation f k * [n] is specifically represented as:

[0020]

[0021] wherein ε k is the introduced Lagrange multiplier, V is the Lyapunov coefficient, Q k [n] represents the queue length of the UAV k in the time slot n, γ k,m [n] is the channel coefficient of the UAV k and the base station m in the time slot n, w k,cm represents the weight value of the energy consumed by the UAV k communication in the time slot n, c represents the number of calculation periods required for each calculation of 1 bit of data, F k is the maximum calculation frequency of the UAV k.

[0022] A computing, communication and control system of a networked UAV, specifically comprising, an initialization unit, a processing unit, a pairing unit, an allocation unit, an optimization unit, an iteration judgment unit, an output unit, an optimization judgment unit and an end unit; the initialization unit is used for initializing state information; the processing unit is used for queue processing; the pairing unit is used for UAV-base station pairing; the allocation unit is used for communication and calculation resource allocation; the optimization unit is used for UAV trajectory optimization; the iteration judgment unit is used for judging whether convergence or the number of iterations reaches the maximum number of iterations; if the preset accuracy is not converged or the number of iterations reaches the maximum number of iterations, the number of iterations is added by 1, and the queue processing is performed again; the output unit is used for outputting the current time slot result if the preset accuracy is converged or the number of iterations reaches the maximum number of iterations; the optimization judgment unit is used for judging whether all time slots are optimized; if all time slots are not optimized, the number of time slots is added by 1, and the initialization of the state information is performed again; the end unit is used for optimization completion if all time slots are optimized, and the process is ended.

[0023] The present application has the following beneficial effects:

[0024] The present application breaks through the drawbacks of the traditional networked UAV mobile edge computing system offline optimization method, and proposes an online optimization method. A computing, communication and control method for networked UAVs and a system implementation thereof are proposed. BRIEF DESCRIPTION OF DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiments or prior art description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments described in the present application, and other drawings can also be obtained by those skilled in the art according to these drawings.

[0026] Figure 1 is the internal structure diagram of the computing, communication and control system of the networked UAV provided by the embodiments of the present application;

[0027] Figure 2A flowchart of a computing, communication and control method for a networked unmanned aerial vehicle is provided according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0029] The present application provides a computing, communication and control method and system for a networked unmanned aerial vehicle, which can effectively solve the randomness problem of a mobile edge computing network. Through the online method, the random joint optimization problem of unmanned aerial vehicle-base station pairing, unmanned aerial vehicle transmission power optimization, bandwidth allocation, computing resource allocation and unmanned aerial vehicle trajectory in the networked unmanned aerial vehicle mobile edge computing system can be solved, so that the system average weighted energy consumption is minimized.

[0030] In a cellular network, multiple unmanned aerial vehicles randomly generate computing tasks in each time slot when performing tasks. In order to alleviate the computing pressure of the unmanned aerial vehicles, the unmanned aerial vehicles select to unload the computing tasks to the ground base stations for help. Considering the random joint optimization problem of unmanned aerial vehicle-base station pairing, unmanned aerial vehicle transmission power optimization, bandwidth allocation, computing resource allocation and unmanned aerial vehicle trajectory in this process, the target is to minimize the system average weighted energy consumption.

[0031] Embodiment one

[0032] As shown in Figure 1 , the present application provides a computing, communication and control system for a networked unmanned aerial vehicle.

[0033] Scenario assumption: assuming M ground base stations and K unmanned aerial vehicles, using set K={1, 2,..., K} to represent all unmanned aerial vehicles, using M={1, 2,..., M} to represent all ground base stations, and the total time slot of the system is N; using λ k,m [n] to represent that unmanned aerial vehicle k establishes a communication connection (pairing) with base station m in time slot n, specifically, λ k,m [n]=1 represents that unmanned aerial vehicle k is paired with base station m in time slot n; otherwise, λ k,m [n]=0; the trajectory of unmanned aerial vehicle k in time slot n is represented as u k [n]; the transmission power of the unmanned aerial vehicle in each time slot is represented as p k [n]; the communication bandwidth of the system in time slot n allocated for task offloading of unmanned aerial vehicle k to base station m is represented as B k,m [n]; and the white noise power spectral density at the base station end is N0.

[0034] E k,cm [n], E k,cp [n], E f [n] respectively represent energy consumed by UAV k communication in time slot n, energy consumed by UAV k computation in time slot n, and energy consumed by UAV k flight in time slot n, and w k,cm , w k,cp , and w f respectively represent weight values of the three, thereby defining system average weighted energy consumption, i.e. objective function, as follows:

[0035]

[0036] wherein E[·] represents mathematical expectation.

[0037] The system of the present application specifically comprises: an initialization unit 110, a processing unit 120, a pairing unit 130, an allocation unit 140, an optimization unit 150, an iteration judgment unit 160, an output unit 170, an optimization judgment unit 180, and an ending unit 190.

[0038] The initialization unit 110 is configured to initialize state information.

[0039] The processing unit 120 is connected with the initialization unit 110 and configured to perform queue processing.

[0040] The pairing unit 130 is connected with the processing unit 120 and configured to perform UAV-base station pairing.

[0041] The allocation unit 140 is connected with the pairing unit 130 and configured to perform communication and computation resource allocation.

[0042] The optimization unit 150 is connected with the allocation unit 140 and configured to perform UAV trajectory optimization.

[0043] The iteration judgment unit 160 is connected with the optimization unit 150 and the processing unit 120 respectively and configured to judge whether convergence is achieved or the number of iterations reaches a maximum number of iterations. If convergence is not achieved to a preset precision or the number of iterations reaches the maximum number of iterations, the number of iterations is increased by 1, and queue processing is performed again.

[0044] The output unit 170 is connected with the iteration judgment unit 160 and configured to output current time slot results if convergence is achieved to the preset precision or the number of iterations reaches the maximum number of iterations.

[0045] The optimization judgment unit 180 is connected with the output unit 170 and the initialization unit 110 respectively and configured to judge whether optimization of all time slots is completed. If optimization of all time slots is not completed, the number of time slots is increased by 1, and initialization of state information is performed again.

[0046] The ending unit 190 is connected with the optimization judging unit 180, and is used for ending the flow if all time slots are optimized.

[0047] Embodiment two

[0048] As Figure 2 shown, a computing, communication and control method for a networked unmanned aerial vehicle is provided, and specifically includes the following steps:

[0049] Step S210: initializing state information.

[0050] The initialization state information includes initializing the flight trajectory of the unmanned aerial vehicle, the maximum transmission power P k of the unmanned aerial vehicle, the total number of time slots, setting n=1, the maximum CPU computing frequency of the unmanned aerial vehicle, the system bandwidth B, the time slot length δ t , and the convergence precision and the maximum number of iterations.

[0051] Specifically, according to the position distribution of the ground user, a straight line trajectory between the starting point and the ending point can be used to initialize the trajectory of the unmanned aerial vehicle.

[0052] Step S220: responding to the initialization state information, performing queue processing.

[0053] The change of the queue length Q k [n+1] of the unmanned aerial vehicle k in two time slots in the system can be represented as:

[0054]

[0055] wherein Q k [n] represents the queue length of the unmanned aerial vehicle k in the time slot n, A k [n] represents the amount of computing tasks generated by the unmanned aerial vehicle k in the time slot n, represents the amount of tasks calculated by the unmanned aerial vehicle k in the time slot n, represents the amount of received tasks in one time slot of the base station m, f k [n] represents the allocated computing resource, and c represents the number of computing periods required for calculating 1 bit of data.

[0056] The amount of received tasks in one time slot of the base station m is represented as:

[0057]

[0058] wherein δ t represents the length of one time slot, γ k,m [n] = β0(||u k [n]-w m || 2 +H​2 -α is the channel coefficient of UAV k and base station m in time slot n, β0is the reference channel power gain, a is the path loss exponent, u k is the trajectory of UAV k in time slot n, H m is the location of base station m, ||.|| is the Euclidean norm.

[0059] Further, the change of the queue length of base station m, Z k,m [n+1] can be expressed as:

[0060]

[0061] where Z k,m [n] is the queue length of base station m in time slot n, D k,m [n] = δ t f k,m [n] / c is the amount of computation that base station m helps UAV k to compute, δ t is the length of a time slot, f k,m [n] is the computation resource allocated by base station m, c is the number of computation cycles needed for computing 1 bit of data, is the amount of received computation of base station m in a time slot.

[0062] Specifically, the queue processing is specifically to process the computation queue at the UAV and the base station respectively by using the Lyapunov method. By using the Lyapunov method to process the change of the computation queue length at the UAV and the base station, the randomness of the system can be eliminated, so that a new objective function E′ sum is obtained:

[0063]

[0064] where V is the Lyapunov coefficient, E sum is the original average weighted energy consumption of the system, Q k [n] is the queue length of UAV k in time slot n, M is the M ground base stations, K is the K UAVs, is the amount of computation of UAV k in time slot n, is the amount of received computation of base station m in a time slot, Z k,m [n] is the queue length of base station m in time slot n, D k,m [n] is the amount of computation that base station m helps UAV k to compute.

[0065] After the queue is processed by using the Lyapunov method, the original random optimization problem becomes a deterministic problem in a discrete time slot.

[0066] ​Step S230: In response to the completion of queue processing, perform drone-base station pairing.

[0067] To obtain the optimal drone-base station pairing scheme, the following expression is used during the drone-base station pairing process to determine the optimal pairing scheme: This represents the optimal communication pairing established between drone k and base station m in time slot n:

[0068]

[0069] Where m * The optimal base station for pairing with drone k during the drone-base station pairing process is determined by the following formula:

[0070]

[0071] Q k [n] represents the queue length of drone k in time slot n, Z k,m [n] represents the queue length of base station m in time slot n, δ t B represents the length of a time slot. k,m [n] represents the communication bandwidth allocated by the system to UAV k for task offloading to base station m in time slot n, p k [n] represents the transmit power of the UAV in each time slot, γ k,m [n]=β0(||u k [n]-w m || 2 +H 2 ) -α For time slot n UAV k The channel coefficients of base station m, N0 represents the white noise power spectral density at the base station, H represents the flight altitude of the UAV, and w m Let m represent the location of base station m, and ||.|| represent the Euclidean norm.

[0072] Step S240: In response to the completion of drone-base station pairing, communication and computing resources are allocated.

[0073] Specifically, the allocation of communication and computing resources involves obtaining the optimal UAV transmission power, optimal UAV bandwidth allocation, and optimal UAV computing resource allocation through the Lagrange duality method.

[0074] Furthermore, for any real number x, it is defined that [x] + =max(0,x),

[0075] Specifically, the optimal allocation of UAV computing resources f k *[n] is specifically expressed as:

[0076]

[0077] wherein ε k is the introduced Lagrange multiplier, V is the Lyapunov coefficient, Q k [n] represents the queue length of the UAV k at the time slot n, γ k,m [n] is the channel coefficient of the UAV k and the base station m at the time slot n, w k,cm represents the weight value of the energy consumed by the UAV k in communication at the time slot n, c represents the number of calculation periods required for calculating 1 bit of data, F k is the maximum calculation frequency of the UAV k.

[0078] At the same time, the Lagrange multiplier ρ k,m , and υ k,m are introduced, and the optimal UAV transmission power P is obtained as:

[0079]

[0080] wherein δ t represents the length of a time slot, γ k,m [n] is the channel coefficient of the UAV k and the base station m at the time slot n, λ k,m [n] represents that the UAV k establishes a communication pair with the base station m at the time slot n, w k,cm represents the weight value of the energy consumed by the UAV k in communication at the time slot n, V is the Lyapunov coefficient, represents the optimal bandwidth allocation, which is determined by the following formula:

[0081]

[0082] Step S250: in response to completing the communication and calculation resource allocation, the UAV trajectory optimization is performed.

[0083] wherein the SCA (Successive Convex Approximation) method is used to update the flight trajectory of the UAV until converging to a stable flight trajectory.

[0084] wherein the step S250 specifically comprises the following sub-steps:

[0085] Step S2501: obtaining input information.

[0086] Specifically, the input information includes the initialization state information in step S210, the processing results of the computing queues at the UAV and the base station in step S220, the UAV-base station pairing results in step S230, and the communication and computing resource allocation results in step S240 as the input information.

[0087] Step S2502: Based on the input information, the optimal UAV flight trajectory is solved and output by using the SCA method.

[0088] Step S260: It is judged whether convergence is reached or the number of iterations reaches the maximum number of iterations.

[0089] Wherein, the steps S220-S250 are iterated constantly, and it is judged whether convergence is reached or the number of iterations reaches the maximum number of iterations. If the preset accuracy is reached or the number of iterations reaches the maximum number of iterations, step S270 is executed.

[0090] If the preset accuracy is not reached or the number of iterations reaches the maximum number of iterations, the number of iterations is increased by 1, and steps S220-S260 are executed again until the preset accuracy is reached or the number of iterations reaches the maximum number of iterations.

[0091] Step S270: The current time slot result is output.

[0092] Wherein, the UAV-base station pairing results, the communication and computing resource allocation results, and the UAV trajectory optimization results are output.

[0093] Step S280: It is judged whether all time slots are optimized.

[0094] Through the constant iteration of steps S220-260 until the preset convergence accuracy or the maximum number of iterations is reached, the optimal result under the current time slot can be obtained, and then the next time slot is entered, and the above steps are repeated until all time slots are optimized.

[0095] If all time slots are not optimized, the number of time slots is increased by 1, and steps S210-S280 are executed again until all time slots are optimized.

[0096] If all time slots are optimized, step S290 is executed: optimization is completed, and the process is ended.

[0097] The present application has the following beneficial effects:

[0098] The present application breaks through the disadvantages of the traditional off-line optimization method of the networked UAV mobile edge computing system, and proposes an online optimization method. A computing, communication and control method for networked UAVs and a system implementation thereof are proposed.

[0099] While the examples of the present application are described with reference to the drawings, they are merely examples and are not intended to limit the present application. Changes, additions and / or deletions can be made to the embodiments without departing from the scope of the present application.

[0100] The above description is only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for computing, communicating and controlling a connected unmanned aerial vehicle, characterized in that, Specifically comprising the following steps: initializing state information; performing queue processing in response to the initialized state information; performing UAV-base station pairing in response to completion of the queue processing; performing communication and computing resource allocation in response to completion of the UAV-base station pairing; performing UAV trajectory optimization in response to completion of the communication and computing resource allocation; judging whether to converge to a preset precision or an iteration number reaches a maximum iteration number in response to completion of the UAV trajectory optimization; outputting a current time slot result if the preset precision is converged to or the iteration number reaches the maximum iteration number; judging whether to optimize all time slots in response to outputting the current time slot result; the process ends if all time slots are optimized; In the process of drone-base station pairing, the best pairing scheme is obtained by the following expression, represents the best communication pairing of drone k with base station m at time slot n: m represents the base station, m * represents the best base station to pair with drone k during the drone-base station pairing process, where Q k [n] represents the queue length of drone k at time slot n, Z k,m [n] represents the queue length of data received from drone k at time slot n by base station m, k represents drone k, δ t represents the length of a time slot, B k,m [n] represents the communication bandwidth allocated by the system to drone k for task offloading to base station m at time slot n, p k [n] represents the transmission power of the drone in each time slot, γ k,m [n] is the channel coefficient of drone k and base station m at time slot n, N0 represents the white noise power spectral density at the base station end. 2.The method of claim 1, wherein, The initialization state information includes an initialization unmanned aerial vehicle flight trajectory, an unmanned aerial vehicle maximum transmission power P k , a total time slot number, an unmanned aerial vehicle maximum CPU calculation frequency, a system bandwidth B, a time slot length δ t , and a convergence precision and a maximum iteration number. 3.The method of claim 1, wherein, performing queue processing includes processing computing queues at the UAV and the base station respectively. 4.The method of claim 3, wherein, where Qkis the queue length at the drone k k [n+1] is represented as: where Q k [n] represents the queue length of UAV k at time slot n, M represents the number of ground base stations, A k [n] represents the amount of computing tasks generated by UAV k at time slot n, represents the amount of computing tasks calculated by UAV k at time slot n, f k [n] represents the allocated computing resources, c represents the number of computing cycles required for each computing 1 bit of data, δ t represents the length of a time slot, represents the amount of tasks transmitted by UAV k to base station m at time slot n, k represents UAV k, and n represents time slot n. 5.The method of claim 4, wherein, where Rm is the amount of reception tasks of base station m in one time slot is represented as: where δ t denotes the length of a time slot, γ k,m [n] = β0(||u k [n] - w m || 2 + H 2 ) -α is the channel coefficient between UAV k and base station m at time slot n, β0is the reference channel power gain, α is the path loss exponent, u k [n] denotes the trajectory of UAV k at time slot n, H denotes the flight altitude of the UAV, w m denotes the location of base station m, ||.|| denotes the Euclidean norm, λ k,m [n] denotes that UAV k establishes a communication connection with base station m at time slot n, B k,m [n] denotes the communication bandwidth allocated by the system to UAV k for task offloading to base station m at time slot n, p k [n] denotes the transmit power of the UAV at each time slot, N0denotes the white noise power spectral density at the base station end. 6.The method of claim 4, wherein, Change Z of the queue length of the base station m k,m [n+1] is expressed as: where Z k,m [n] represents the data queue length received by base station m from UAV k at time slot n, D k,m [n] = δ t f k,m [n] / c represents the amount of computation that base station m helps UAV k to calculate, δ t represents the length of a time slot, f k,m [n] represents the computation resource allocated by base station m for UAV k at time slot n, c represents the number of computation cycles required for calculating 1 bit of data, k represents UAV k, and n represents time slot n.

7. The method of claim 1, wherein the method further comprises: The communication and computing resource allocation includes obtaining an optimal drone computing resource allocation, the optimal drone computing resource allocation f k * [n] is specifically represented as: where ε k is the introduced Lagrange multiplier, V is the Lyapunov coefficient, Q k [n] represents the queue length of the UAV k at the time slot n, γ k,m [n] is the channel coefficient of the UAV k and the base station m at the time slot n, w k,cm represents the weight value of the energy consumed by the UAV k communication at the time slot n, c represents the number of calculation periods required for each calculation of 1 bit of data, F k is the maximum calculation frequency of the UAV k.

8. A computing, communication and control system for a connected drone, characterized in that, Specifically comprising an initialization unit, a processing unit, a pairing unit, an allocation unit, an optimization unit, an iteration judging unit, an output unit, an optimization judging unit, and an ending unit; the initialization unit is configured to initialize state information; the processing unit is configured to perform queue processing; the pairing unit is configured to perform UAV-base station pairing; the allocation unit is configured to perform communication and computing resource allocation; the optimization unit is configured to perform UAV trajectory optimization; the iteration judging unit is configured to judge whether to converge or an iteration number reaches a maximum iteration number; if the preset precision is not converged to or the iteration number reaches the maximum iteration number, the iteration number is increased by 1, and the queue processing is performed again; the output unit is configured to output a current time slot result if the preset precision is converged to or the iteration number reaches the maximum iteration number; the optimization judging unit is configured to judge whether to optimize all time slots; if all time slots are not optimized, the time slot number is increased by 1, and the initialization of the state information is performed again; the ending unit is configured to optimize if all time slots are optimized, and the process ends; The pairing unit, in the process of drone-base station pairing, obtains the best pairing scheme by the following expression, represents the best communication pairing of drone k with base station m established at time slot n: m represents the base station, m * represents the best base station to pair with drone k during the drone-base station pairing process, where Q k [n] represents the queue length of drone k at time slot n, Z k,m [n] represents the queue length of data received by base station m from drone k at time slot n, k represents drone k, δ t represents the length of a time slot, B k,m [n] represents the communication bandwidth allocated by the system to drone k for task offloading to base station m at time slot n, p k [n] represents the transmission power of the drone in each time slot, γ k,m [n] is the channel coefficient of drone k and base station m at time slot n, N0 represents the white noise power spectral density at the base station end.

Images