STAR-RIS-assisted UAV-MEC network joint optimization method

Abstract

The invention discloses an STAR-RIS-assisted UAV-MEC network joint optimization method, which is characterized in that STAR-RIS is applied to a scene of a UAV-assisted MEC wireless communication network, and three sub-problems of task unloading decision, UAV hovering position and resource allocation are jointly optimized so as to minimize the total energy consumption of a system. According to the embodiment of the invention, the hovering unmanned aerial vehicle equipped with the MEC server and the STAR-RIS is adopted as an air BS to cooperate with a ground BS to provide calculation and communication services for users, and the position of the unmanned aerial vehicle, the amplitude of the STAR-RIS, the phase shift and the task unloading ratio are jointly optimized. According to the present invention, the energy consumption of a system in an energy splitting (ES) mode is minimized.

Description

Technical Field

[0001] The present invention relates to the field of communications, and in particular to a STAR-RIS-assisted UAV-MEC network joint optimization method. Background Art

[0002] The emergence of the sixth generation of communication technology (6G) enables smart devices to provide ultra-high-speed and ultra-low latency services to mobile terminals (MTs), such as extended reality and 8K ultra-high-definition video transmission. Cloud computing, as a centralized computing paradigm, can meet users' Quality of Service (QoS) by offloading tasks to data centers. However, the long propagation distance from cloud computing to MTs is not suitable for computationally intensive and latency-sensitive applications. Compared with cloud computing, multi-access edge computing (MEC) can quickly complete tasks offloaded from MTs by deploying computing and caching resources at the edge of the network, such as base stations (BSs). However, the limited signal coverage of traditional BSs makes it challenging to establish a reliable connection between mobile MTs and BSs.

[0003] Unmanned aerial vehicle (UAV) assisted communication network has become one of the most promising architectures in the field of task offloading in recent years due to its advantages such as flexible deployment, low cost-effectiveness and higher line-of-sight (LoS) communication probability. UAVs equipped with MEC servers not only have the computing power to improve the task response speed, but also can act as relays to transmit tasks to other BSs for processing. Given the limited onboard energy of UAVs, it is urgent to explore energy-saving solutions to improve the performance of UAV-assisted MEC systems.

[0004] Equipped with low-power and low-cost metasurfaces, reconfigurable intelligent surfaces (RIS) can enhance channel conditions and signal coverage by dynamically adjusting the wireless propagation environment, thereby improving the communication capacity and energy efficiency of UAV-assisted MEC systems. In RIS-assisted UAV communication networks, the moving UAV may cause the signal source and destination to be located on both sides of the RIS. However, in RIS, the incident signal can only be reflected, which requires that the transmitter and receiver must be located on the same side. Therefore, in order to provide a full-space radio environment, a simultaneous transmitting and reflecting RIS (STAR-RIS) has been recently proposed. The STAR-RIS-assisted UAV network can break the geographical restrictions of access terminals and improve the coverage and communication quality of wireless communications. In the STAR-RIS-assisted UAV-MEC network, it is necessary to comprehensively consider the mobility of UAVs, the coefficient matrix of STAR-RIS, and the offloading strategy to achieve the minimum system energy consumption. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, the present invention provides a TAR-RIS-assisted UAV-MEC network joint optimization method.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: A STAR-RIS-assisted UAV-MEC network joint optimization method comprises the following steps: S1. Build a system model based on the task offloading process in the STAR-RIS-assisted UAV-MEC network model and determine the objective function and constraints of the system model, including the network model, communication model and energy consumption model; S2, using KKT condition optimization solution method to solve the optimal task offloading decision in the task offloading process; S3, using continuous convex approximation method to optimize the hovering position of the UAV; S4, using continuous convex approximation method to optimize the resource allocation of UAVs; S5. Alternately solve the unloading ratio, the hovering position of the UAV, and the resource allocation of the system model in the unloading process, and set a threshold to make it converge, so as to obtain the optimal solution of the task unloading ratio, the hovering position of the UAV, and the resource allocation of the system model.

[0007] Furthermore, the S1 specifically includes the following steps: S11. Build a STAR-RIS-assisted UAV-MEC network model, including a single antenna , drones equipped with MEC servers and STAR-RIS , and two single-antenna users and users , where each column and row of STAR-RIS consists of and Passive unit, The components represent the Column and Elements of the row; S12, build a communication model, establish a Cartesian coordinate system, and user and single antenna The positions of and Description, drone u Hovering in a fixed position to serve the user, drone position and STAR-RIS elements The locations are and ; S13, build energy consumption model, user The transmission energy consumption and nodes The computational energy consumption is expressed as:

[0008]

[0009] In the formula, is the transmission energy consumption of user i, is the computational energy consumption of node k, is the transmission power of user i, represents the capacitance coefficient, For Node CPU frequency, For users Upload the task to the node Uplink transmission time, For Node Counting users The duration of the task; S14. Determine the objective function and constraints of the system model.

[0010] Furthermore, the objective function and constraint conditions in S14 are expressed as:

[0011]

[0012]

[0013]

[0014]

[0015]

[0016] In the formula, is the total energy of the STAR-RIS-assisted UAV-MEC communication network, is the transmission or reflection coefficient matrix of STAR-RIS, is the drone location, is the task offloading ratio, For users The task offloading ratio is To process users The computing resources required for the task, , are the capacitance coefficients of the uth and bth nodes respectively, is the CPU frequency of the b-th node; , TAR-RIS The magnitude and phase of each element.

[0017] Furthermore, the S2 specifically includes the following steps: S21, fix the UAV position, transmission and reflection coefficient matrices and rewrite the objective function and constraints; S22, introduce the approximate quadratic regularization term to optimize the transmission or reflection coefficient matrix of STAR-RIS and set the drone position The non-convexity problem arising from S23. Use KKT conditions to solve the optimal task offloading decision in the task offloading process.

[0018] Furthermore, the optimal task offloading decision in the offloading process in S23 is expressed as:

[0019] In the formula, Make decisions for the most effective task offloading, is the total energy consumption when the UAV position, transmission and reflection coefficient matrices are fixed, For fixed position, are the transmission and reflection coefficient matrices, is the offload task ratio, is a non-negative Lagrange multiplier.

[0020] Furthermore, the S3 specifically includes the following steps: S31, given task offloading ratio and the transmission and reflection coefficient matrices and introduce approximate variables; S32, use the scaling method to solve the lower bound of the non-convex constraint, and use the Lagrangian dual decomposition method to solve the hovering position of the drone .

[0021] Furthermore, the hovering position of the drone solved in S32 is expressed as:

[0022] In the formula, To solve the optimal drone hovering position, is the hovering position of the drone, is a non-negative Lagrange multiplier, is the Lagrangian function.

[0023] Furthermore, the S4 specifically includes the following steps: S41, set the fixed task offloading ratio and the drone hovering position, and rewrite the objective function and constraint conditions; S42, introduce approximate variables to optimize the non-convex problem under the conditions of fixed task offloading ratio and drone hovering position; S43. Use Lagrangian dual decomposition method to solve the fixed transmission and reflection coefficient matrices to obtain the optimal resource allocation strategy.

[0024] Furthermore, the optimal resource allocation strategy in S3 is expressed as:

[0025] In the formula, is the optimal resource allocation strategy, is the resource allocation status of user i, is the Lagrange multiplier, is the dual Lagrangian function.

[0026] The present invention has the following beneficial effects: (1) This paper proposes a new task offloading, UAV hovering position and resource allocation problem to minimize the energy consumption of STAR-RIS assisted UAV-MEC wireless network. This is the first time that energy-efficient task offloading and resource allocation for UAVs equipped with both MEC servers and STAR-RIS are studied in UAV communication networks.

[0027] (2) The present invention decomposes the energy minimization problem into the task offloading decision subproblem, the UAV hovering position subproblem and the transmission resource allocation subproblem. For the task offloading decision subproblem, the KKT condition is used to optimize the task offloading ratio. For the UAV hovering position subproblem, SCA is used to optimize the UAV position. For the transmission resource allocation subproblem, SCA is used to optimize the STAR-RIS beamforming.

[0028] (3) Simulation results verify that compared with the existing (STAR-)RIS-assisted UAV-MEC wireless network, this approach effectively reduces energy consumption and achieves fast convergence in a heterogeneous simulation environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a UAV-MEC wireless communication scenario diagram assisted by STAR-RIS of the present invention; Figure 2 It is the time allocation diagram for task execution of the system of the present invention; Figure 3 It is a framework diagram for solving the system of the present invention; Figure 4 It is a simulation diagram of the convergence performance of the algorithm of the present invention; Figure 5 This is a simulation diagram of the effect of unloading ratio and UAV position on performance of the present invention; Figure 6 This is a simulation diagram of the impact of network bandwidth on performance in the present invention; Figure 7 This is a simulation diagram of the impact of channel noise on performance of the present invention; Figure 8 This is a simulation diagram of the effect of user transmission power on performance of the present invention; Fig. 9 This is a simulation diagram showing the effect of the number of smart reflective surface elements on performance of the present invention; Fig.10 This is a simulation diagram of the impact of the task computing requirements on performance of the present invention. DETAILED DESCRIPTION

[0030] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0031] A STAR-RIS-assisted UAV-MEC network joint optimization method, such as Figure 1 As shown, the following steps are included: S1. Build a system model based on the task offloading process in the STAR-RIS-assisted UAV-MEC network model and determine the objective function and constraints of the system model, including the network model, communication model and energy consumption model; In this embodiment, the following steps are specifically included: S11. Analyze and build a STAR-RIS-assisted UAV-MEC network model The STAR-RIS-assisted UAV-MEC access network proposed in the present invention is as follows: Figure 1 As shown. It includes a single antenna BS 、A UAV with MEC server and STAR-RIS installed , and two single-antenna users. Specifically, each column and row of STAR-RIS consists of and Passive units. The components represent the Column and The two users in the transmission space and reflection space are called users and users This embodiment assumes that the direct communication link between the two users and the ground base station is blocked, and each STAR-RIS element operates in Energy Splitting (ES) mode. It can partially process tasks uploaded by users and return the results to users, while the rest of the tasks are offloaded to the BS through STAR-RIS components Since the connection between MEC and UAV uses low-latency, interference-free, and high-capacity optical fiber, the data transmission delay of the link between UAV and MEC server is not considered in this process. In addition, considering that UAV and BS often have higher transmission power, the calculation result is usually small, and this embodiment ignores the delay in the local execution and task offloading stages.

[0032] The present invention considers a partial offloading scheme for delay-sensitive computational tasks in a STAR-RIS-assisted UAV-MEC system. This computation offloading model allows and BS Parallel processing tasks on UAV The tasks processed on the ground are called local tasks, while the tasks offloaded to the ground BS are called local tasks. The task is called an offload task. Figure 2 As shown, the time allocation of task processing in a STAR-RIS-assisted UAV-MEC system is demonstrated, where two users utilize a time-dependent In the local execution phase, the tasks uploaded by the user are executed on the UAV with computing power. During the task offloading phase, the remaining tasks are forwarded to the ground BS via STAR-RIS. When the mission is completed, the UAV and BS The calculation results are returned to the user. and BS They often have higher transmission power and the calculation results are usually small, so the download time is relatively negligible during the local execution phase and task offloading phase.

[0033] S12. Analyze and build communication model In order to describe the positions of the two users, the UAV, STAR-RIS and the ground BS, this embodiment establishes a three-dimensional Cartesian coordinate system. and BS The positions of and vector This embodiment assumes that the UAV Hovering in a fixed position to serve the user. UAV Location and STAR-RIS Components The locations are and .

[0034] Due to the high probability of LoS links in UAV communications, this embodiment assumes that the user With UAV Communication channels between users With communication channels between STAR-RIS elements, and STAR-RIS components and BS The communication channels between nodes are all LoS links and follow the free space path loss model. and The channel gain between can be expressed as: (1) in, , , , is the transmit power distance The received power, and . By node To Node The signal-to-interference-plus-Noise Ratio (SINR) of the wireless link is calculated using Indicates, expressed as (2) where is the transmission power, From user to UAV or BS The variance of the noise power in the link is To Node The transmission rate is expressed as (3) in Represents a slave node To Node channel transmission bandwidth.

[0035] Channels where tasks are executed locally UAV Equipped with computing resources, allowing users Upload some missions to the UAV Processing. UAV The received signal can be written as (4) in For users signal, It is UAV Received additive white Gaussian noise.

[0036] According to equations (2) and (3), from the user To UAV The signal-to-noise ratio and transmission rate are respectively and express.

[0037] Channel for task offloading Due to the limitation of UAV energy and mission time, UAV Can only perform part of the user's tasks. BS STAR-RIS handles the rest of the tasks. To STAR-RIS components The channel gain is from STAR-RIS Components to BS The channel gain is expressed as Therefore, from the user to STAR-RIS and from STAR-RIS to BS The channel gains are expressed as and .

[0038] From the user The STAR-RIS transmission or reflection coefficient matrix of the incident signal, where STAR-RIS The amplitude of each element , Phase Therefore, when STAR-RIS operates in ES mode, there are the following constraints: (5) Therefore, BS The signal received at (6) in It's BS Received additive white Gaussian noise.

[0039] Similarly, according to equations (2) and (3), from the user BS via STAR-RIS The signal-to-noise ratio and transmission rate are and express.

[0040] S13. Analyze and build energy consumption model when and , indicating that the user Upload local missions to the UAV Processing. and , indicating that the user Upload the remaining tasks to BS to be processed. Two users upload tasks to the node The maximum value of the uplink transmission time is ,in Is a user Upload the task to the node Uplink transmission time. Represented as a node Counting users Then, and Respectively expressed as (7) (8) If , ;otherwise, , . For users The task offloading ratio. To process users The computing resources required for the task, that is, the number of CPU cycles required to complete 1 bit of input data; Is a node The CPU frequency (cycles per second).

[0041] Therefore, during the task execution, the user The transmission energy consumption and nodes The computational energy consumption is (9) (10) in represents the capacitance coefficient, which is affected by the node The influence of the processor's chip architecture.

[0042] S14. Determine the optimization objective function and constraints From equations (9) and (10), we can see that the total energy consumption of the STAR-RIS-assisted UAV-MEC communication network proposed in this embodiment is indicates that, and , ,but (11) According to formula (11), the energy minimization problem of the STAR-RIS assisted UAV-MEC system can be expressed as (12a) (12b) (12c) (12d) (12e) (12f) (12g) Constraint (12b) indicates that the task offloading ratio of each user is between 0 and 1. Constraint (12c) indicates that in UAV Tasks processed on the Constraint (12d) requires that BS The tasks to be processed must be within the allocated offload interval The constraints (12e) and (12f) are about the STAR-RIS The constraints (12g) specify the amplitude and phase shift requirements of the UAV The hover position range.

[0043] The flying UAV can change the channel gain between the user and the UAV and between the user and the STAR-RIS-base station, thereby affecting the transmission energy consumption. of and is fixed, and the UAV Same location as STAR-RIS. of Optimize to minimize system energy consumption.

[0044] like Figure 3 As shown, this embodiment uses an alternating optimization technique to decompose problem (12) into three sub-problems: Task offloading decision subproblem: Given , , that is, when and Fixed issue (12) when optimizing UAV offloading decisions , to achieve the minimum energy consumption. This embodiment uses KKT condition to optimize task offloading decision, using express.

[0045] Task offloading decision subproblem: Given , , Problem (12) Optimizing UAV Hover position , to achieve the minimum energy consumption. This embodiment uses the SCA method to optimize the hovering position. express.

[0046] STAR-RIS resource allocation subproblem: Given and , transmission and reflection coefficient matrices is optimized to minimize the transmission energy of the user offloading task. This embodiment uses the SCA method to obtain an optimal solution. express.

[0047] S2, using KKT condition optimization solution method to solve the optimal task offloading decision in the task offloading process; First given , solve the problem (12). When , the goal of problem (12) is to make the optimal unloading decision and the optimal hovering position of the UAV to minimize the energy consumption.

[0048] S21. Fix the UAV position and transmission and reflection coefficient matrices, and rewrite the objective function and constraints. have .

[0049] In this embodiment, the objective function can be rewritten as (13a)

[0050] (13b) in Defined as . Fixed UAV hover position , Problem (13) can be rewritten as (14a)

[0051] (14b) (14c) S22, introduce the approximate quadratic regularization term Considering the given and Under the non-convexity of problem (14) generated by the optimization objective (14a), a near-quadratic regularization term is added, namely To overcome this problem, is a positive scalar parameter. The regularization problem is is equivalent to the original problem. So (15) Problem (14) can be written as (16)

[0052] S23. Obtain a feasible solution set through KKT conditions.

[0053] According to (16), is a The Hessian matrix of problem (16) is positive semidefinite, so problem (16) can be classified as a convex optimization problem. The optimal value can be obtained by the KKT condition. The Lagrangian function of problem (16) is It can be expressed as (17), where , and are the non-negative Lagrange multipliers corresponding to constraints (12b), (14b) and (14c), respectively.

[0054] (17) Optimal task offloading decision based on KKT conditions It can be expressed as (18) if is a feasible solution set At any point that satisfies the following KKT conditions, the optimal task offloading decision can be obtained, as shown in (19) (19a) (19b) (19c) (19d) (19e) (19f) (19g) (19h) (19i) (19j) S3, based on the optimal solution obtained in S2, the continuous convex approximation method is used to optimize the hovering position of the UAV and the resource allocation of the system model; In this part, when When , the goal of problem (13) is to determine the optimal UAV hovering position with the goal of minimizing energy consumption.

[0055] S31, given task offloading ratio and the transmission and reflection coefficient matrices , introducing approximate variables The non-convexity of objective (13a) stems from and , the constant term and its given initial value Replace them with their approximate variables, respectively and It means that by transforming (13a) into a convex function, problem (13) can be reformulated as:

[0056]

[0057] (20a) (20b) (20c) (20d) S32. Using the scaling method to derive the lower bound of non-convex constraints In order to solve the non-convexity of (20c) and (20d), this example uses the SCA method to derive a near-optimal solution. , as shown in (21) (twenty one) Among them It's very tight. and About The two approximate constants of are defined as follows (twenty two) (twenty three) Then , constraint (20c) can be approximated by its concave lower bound, (twenty four) in Expressed as (25) in .

[0058] Similarly, constraint (20d) can be approximated by its concave lower bound given by (26) Therefore, problem (20) can be approximated by a convex problem as follows

[0059]

[0060] (27a) (27b) (27c) (27d) S33, using Lagrangian dual decomposition method to solve the UAV hovering position

[0061] CVX can be used to solve the optimal hovering position of the drone This embodiment uses the Lagrangian dual method to improve the solution efficiency. Lagrangian function It is given in (28), where , , , and is the Lagrange multiplier corresponding to the constraint in (28).

[0062] (28) The Lagrangian dual function of (28) can be given by (29) Using the convexity of problem (27), the optimal solutions of the original problem (27) and the dual problem (29) both satisfy the KKT condition. , and get the optimal solution Optimal hovering position ,in From (30) we can get (30) Lagrange multipliers Update by (31) (32) (33) (34) (35) In the above formula It is in Update the Lagrange multiplier at iteration The step length. Updated in each iteration. Without loss of generality, this embodiment assumes Each element of has the same step size during the iteration.

[0063] S4, using continuous convex approximation method to optimize the resource allocation of UAVs; In this part, when , When , the goal of problem (12) is to determine the optimal STAR-RIS resource allocation with the goal of minimizing energy consumption.

[0064] S41, given task offloading ratio and UAV hovering position , rewrite the objective function and constraints Therefore, problem (12) can be rewritten as (36a)

[0065] (36b) in .

[0066] S42, introduce approximate variables and use the scaling method to derive the lower bound of non-convex constraints. For a given and When , the non-convexity of the objective (36a) and constraint (36b) comes from . Similarly, an approximate variable replace to overcome its non-convexity. Problem (36) is rewritten as (37a)

[0067] (37b) (37c) for , It can be written as (38) in , and Expressed as (39) (40) (41) Therefore, constraints (12e) and (12f) are replaced by the following constraints (42) in yes Middle elements. Obviously, (42) is convex.

[0068] In order to solve the non-convex constraint (37c), the SCA method is adopted. Specifically, since for is convex, so its lower bound can be derived as follows: (43) in, is obtained in the previous iteration. Then, constraint (37c) can be replaced by the following constraint (44) This is linear, so problem (37) can be rewritten as (45)

[0069] S44, use Lagrangian dual decomposition method to solve the fixed transmission and reflection coefficient matrices

[0070] Although the convex problem (45) can be solved, for example, using the CVX toolbox, this embodiment utilizes the Lagrangian dual method to improve computational efficiency. Lagrangian function It is given in (46), where , and are the Lagrange multipliers associated with the three constraints of problem (45).

[0071] (46) The Lagrange dual function is written as (47) Using the convexity of problem (45), the optimal solutions of the original problem (45) and the dual problem (47) both satisfy the KKT condition. , the optimal solution is derived . We obtained the optimal for (48) Lagrange multipliers Update by (49) (50) (51) In the formula It is in Update the Lagrange multiplier at iteration The step vector.

[0072] S5. Alternately solve the unloading ratio, the hovering position of the UAV, and the resource allocation of the system model in the unloading process, and set a threshold to make it converge, so as to obtain the optimal solution of the task unloading ratio, the hovering position of the UAV, and the resource allocation of the system model.

[0073] Use Matlab or other mathematical computing tools to alternately solve the above three sub-problems and set appropriate convergence thresholds until convergence to obtain the optimal solution for task offloading, UAV position and resource allocation.

[0074] Analyze the impact of various factors on performance This step will provide numerical results to demonstrate the performance of the proposed STAR-RIS-assisted UAV-MEC strategy in a dual-user wireless communication system.

[0075] Analyze the algorithm convergence performance under different RIS and STAR-RIS exist Figure 4 In this embodiment, the convergence performance of the proposed algorithm is studied. The results show that this method can ensure that the energy consumption converges to the optimal value after multiple iterations, and determines the optimal task offloading ratio, the hovering position of the UAV, and the transmission and reflection coefficients. In addition, the convergence speed of the three benchmark test schemes is slower than that of the solution proposed in this embodiment. The amplitude of the STAR-RIS / RIS element and the hovering position of the UAV are optimized by the random phase shift method, and the reflection matrix and the hovering position of the UAV are optimized by the RIS method. Although the number of corresponding optimization variables is reduced compared with the proposed solution when the system is working, the number of initial variables is greater than the number of solutions in this embodiment. The setting of the initial variables has a significant impact on the number of iterations to obtain the optimal variables.

[0076] Analyze the impact on performance at different unloading ratios and UAV hovering positions Figure 5 The relationship between the offloading ratio, the hovering position of the UAV and the total energy consumption of the system was explored. It was observed that the total energy consumption increased with the increase of the offloading ratio. The reason is that the larger the offloading ratio, the more the user's computing tasks are forwarded to the ground BS instead of the UAV for processing, resulting in more transmission energy consumption caused by long-distance transmission. In addition, the hovering position of the UAV has a certain impact on the total energy consumption of the system. By optimizing the hovering position of the UAV, computing services can be provided to users. At the same time, the STAR-RIS installed on the UAV can dynamically adjust the transmission signal as the hovering position of the UAV changes, and reflect the signal to the ground BS. Therefore, optimizing the offloading ratio and the hovering position of the UAV can help reduce the total energy consumption of the system.

[0077] Analyze the impact of network bandwidth on performance under different RIS and STAR-RIS exist Figure 6In this embodiment, the total energy consumption of the system under different network bandwidths is plotted. As the network bandwidth increases, the total energy consumption of all solutions decreases. The reason is that the growth of network bandwidth increases the transmission rate between users and UAVs, as well as the transmission rate from users to ground BSs through STAR-RIS / RIS, reducing transmission delay and energy consumption. At the same time, by jointly optimizing the transmission and reflection matrices and task offloading decisions, the proposed solution always outperforms similar solutions, thereby narrowing the gap between all solutions. In addition, a wider network bandwidth can reduce the impact of random phase shifts, improve the non-ideal channel conditions caused by RIS, and consume less transmission energy.

[0078] Step 6.4: Analyze the impact of channel noise on performance under different RIS and STAR-RIS exist Figure 7 In this embodiment, the total energy consumption is plotted when the signal-to-noise ratio changes. A higher signal-to-noise ratio means that the signal strength is greater relative to the noise strength. As the signal-to-noise ratio increases, the total energy consumption of all schemes is reduced due to the improvement of the signal transmission environment and the reduction of transmission energy. At the same time, Figure 7 As shown, the proposed solution always outperforms other solutions and Figure 7 The gap between all solutions is reduced. It can be concluded from this embodiment that the solution proposed in this embodiment has a strong anti-interference performance and is more suitable for complex and challenging communication environments.

[0079] Analyze the impact of user transmission power on performance under different RIS and STAR-RIS This embodiment plots the total energy consumption variation trend when the user transmission power changes. As the user transmission power increases, the transmission energy consumption increases, and the performance of all solutions decreases. Since STAR-RIS introduces additional LoS links for two users to offload their tasks, the proposed solution performs better by jointly optimizing the transmission solution and the reflection matrix, as well as the task offloading decision. In addition, as the user transmission power increases, the gap between STAR-RIS and RIS also increases. This is because STAR-RIS can improve the user's transmission rate in the transmission and reflection areas, while RIS can only improve the user's transmission rate in the reflection area.

[0080] Analyze the impact of the number of RIS / RIS components on performance under different RIS and STAR-RIS exist Fig. 9In this embodiment, the energy consumption of all schemes is studied when the number of STAR-RIS / RIS elements changes. As the number of elements increases, both STAR-RIS and RIS can provide more transmission channels for the incoming event signal, improving the transmission environment and thus reducing the energy of the system. At the same time, the performance of the STAR-RIS scheme is significantly better than the RIS-only scheme, and the gap is widened by jointly optimizing the transmission and reflection matrices. In addition, as the number of elements increases, the gap between the optimal phase shift and the random phase shift also increases. This means that the performance of STAR-RIS and RIS can be improved by optimizing the phase shift of all elements.

[0081] Analyze the impact of mission computing requirements on performance under different RIS and STAR-RIS exist Fig.10 In this embodiment, the effect of different CPU cycles required to calculate 1 bit of task data on the energy consumption of all solutions is studied. As the CPU cycles required to calculate 1 bit of task data increase, the UAV and the ground BS consume more computing energy, which reduces the performance of all solutions. In addition, the solution using STAR-RIS performs better than the solution using only RIS. The reason is that due to the limited computing power of the UAV, the improvement of the transmission channel will offload more tasks to the BS for processing. Similarly, as the computing energy consumption increases, the gap between having the optimal phase shift and the random phase shift will also increase.

[0082] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0083] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0084] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0085] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0086] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A STAR-RIS-assisted UAV-MEC network joint optimization method, characterized in that: The steps include: S1. Build a system model based on the task offloading process in the STAR-RIS-assisted UAV-MEC network model and determine the objective function and constraints of the system model, including the network model, communication model and energy consumption model; S2, using KKT condition optimization solution method to solve the optimal task offloading decision in the task offloading process; S3, using continuous convex approximation method to optimize the hovering position of the UAV; S4, using continuous convex approximation method to optimize the resource allocation of UAVs; S5. Alternately solve the unloading ratio, the hovering position of the UAV, and the resource allocation of the system model in the unloading process, and set a threshold to make it converge, so as to obtain the optimal solution of the task unloading ratio, the hovering position of the UAV, and the resource allocation of the system model.

2. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 1, characterized in that: The S1 specifically includes the following steps: S11. Build a STAR-RIS-assisted UAV-MEC network model, including a single antenna , drones equipped with MEC servers and STAR-RIS , and two single-antenna users and users , where each column and row of STAR-RIS consists of and Passive unit, The components represent the Column and Elements of the row; S12, build a communication model, establish a Cartesian coordinate system, and user and single antenna The positions of and Description, drone u Hovering in a fixed position to serve the user, drone position and STAR-RIS elements The locations are and ; S13, build energy consumption model, user The transmission energy consumption and nodes The computational energy consumption is expressed as: In the formula, is the transmission energy consumption of user i, is the computational energy consumption of node k, is the transmission power of user i, represents the capacitance coefficient, For Node CPU frequency, For users Upload the task to the node Uplink transmission time, For Node Counting users The duration of the task; S14. Determine the objective function and constraints of the system model.

3. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 2, characterized in that: The objective function and constraint conditions in S14 are expressed as: In the formula, is the total energy of the STAR-RIS-assisted UAV-MEC communication network, is the transmission or reflection coefficient matrix of STAR-RIS, is the drone location, is the task offloading ratio, For users The task offloading ratio is To process users The computing resources required for the task, , are the capacitance coefficients of the uth and bth nodes respectively, is the CPU frequency of the b-th node; , TAR-RIS The amplitude and phase of each element, , is a constant term.

4. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 1, characterized in that: The S2 specifically includes the following steps: S21, fix the UAV position, transmission and reflection coefficient matrices and rewrite the objective function and constraints; S22, introduce the approximate quadratic regularization term to optimize the transmission or reflection coefficient matrix of STAR-RIS and set the drone position The non-convexity problem arising from S23. Use KKT conditions to solve the optimal task offloading decision in the task offloading process.

5. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 4, characterized in that ,The optimal task offloading decision in the offloading process in S23 is expressed as: In the formula, Make decisions for the most effective task offloading, is the total energy consumption when the UAV position, transmission and reflection coefficient matrices are fixed, For fixed position, are the transmission and reflection coefficient matrices, is the offloading task ratio, is a non-negative Lagrange multiplier.

6. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 1, characterized in that , the S3 specifically includes the following steps: S31, given task offloading ratio and the transmission and reflection coefficient matrices and introduce approximate variables; S32, use the scaling method to solve the lower bound of the non-convex constraint, and use the Lagrangian dual decomposition method to solve the hovering position of the drone .

7. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 6, characterized in that: The hovering position of the UAV solved in S32 is expressed as: In the formula, To solve the optimal drone hovering position, is the hovering position of the drone, is a non-negative Lagrange multiplier, is the Lagrangian function.

8. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 1, characterized in that: The S4 specifically includes the following steps: S41, set the fixed task offloading ratio and the drone hovering position, and rewrite the objective function and constraint conditions; S42, introduce approximate variables to optimize the non-convex problem under the conditions of fixed task offloading ratio and drone hovering position; S43. Use Lagrangian dual decomposition method to solve the fixed transmission and reflection coefficient matrices to obtain the optimal resource allocation strategy.

9. A STAR-RIS-assisted UAV-MEC network joint optimization method according to claim 8, characterized in that: The optimal resource allocation strategy in S3 is expressed as: In the formula, is the optimal resource allocation strategy, is the resource allocation status of user i, is the Lagrange multiplier, is the dual Lagrangian function.

Images