Modeling and solving method for resource allocation model of underwater optical communication data acquisition system assisted by AUV (Autonomous Underwater Vehicle)

Abstract

The invention discloses a modeling and solving method for a resource allocation model of an AUV (Autonomous Underwater Vehicle)-assisted underwater optical communication data acquisition system. The modeling method comprises the following steps: constructing a wireless communication model of the AUV-assisted UVLC data acquisition system; establishing a channel gain model between the AUV and the USN; establishing a channel gain model between the AUV and the buoy collector; representing the total time delay of data acquisition according to the navigation time of the AUV from the starting position to the ending position and the total time of data transmission of the AUV in the nth time slot; by jointly optimizing multi-dimensional communication resources of a spectrum allocation variable, a time slot division variable and an AUV route planning variable, an AUV-assisted UVLC data acquisition system resource allocation model is established for an optimization problem P1 of minimizing total data acquisition time delay. On one hand, the modeling method which is complete in theory and feasible in engineering is provided for end-to-end time delay optimization of the underwater mobile communication system; furthermore, according to the solving method provided by the invention, the calculation complexity of the non-convex problem can be effectively reduced, and a low-delay optimization scheme which is rigorous in theory and feasible in engineering is provided for an underwater mobile optical communication system.

Description

Technical Field

[0001] The present invention relates to a modeling and solving method for a resource allocation model of an AUV-assisted underwater optical communication data acquisition system, and belongs to the technical field of wireless communication. Background Art

[0002] In recent years, with the increasing demand for marine resource exploration, environmental monitoring and disaster prevention, the necessity of underwater data acquisition has become increasingly prominent. Against this background, the underwater visible light communication (UVLC) data acquisition system assisted by autonomous underwater vehicles (AUVs) has received great attention from academia and industry. Although the mainstream underwater acoustic communication protocol has a coverage range of thousands of kilometers, its limited spectrum resources result in a data transmission rate of only tens of Kbps. Compared with underwater acoustic communication technology, UVLC spectrum resources can achieve a high-speed transmission rate of 100Mbps, but due to the absorption and scattering effects of seawater, the effective communication distance of UVLC is limited to less than 200 meters. The present invention effectively expands the coverage of the UVLC network by introducing AUV as a dynamic communication relay node to meet the communication challenges of complex underwater environments. The core difficulty of system design is that the three-dimensional navigation trajectory of the AUV is closely related to the time-frequency domain resources. Specifically, the navigation of the AUV will change the communication link distance, which will then cause channel gain fluctuations, which can easily cause data backlog and transmission delay.

[0003] Therefore, the present invention constructs an AUV-assisted UVLC data acquisition system. By establishing an AUV-assisted underwater optical communication data acquisition system resource allocation model, the goal of minimizing the total data acquisition delay is ultimately achieved, thus meeting the real-time requirements of underwater services. Summary of the Invention

[0004] The present invention provides a modeling method for a resource allocation model of an AUV-assisted underwater optical communication data acquisition system, aiming to jointly regulate AUV trajectory planning and time-frequency domain communication resource allocation and construct the resource allocation model of the AUV-assisted underwater optical communication data acquisition system by minimizing the total data acquisition delay. The present invention further provides an efficient algorithm to provide a feasible basis for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system, thereby providing an effective technical means for solving the resource allocation problem in underwater data acquisition.

[0005] The technical solution of the present invention is:

[0006] According to a first aspect of the present invention, a modeling method for a resource allocation model of an AUV-assisted underwater optical communication data acquisition system is provided, comprising: constructing a wireless communication model of an AUV-assisted UVLC data acquisition system; the wireless communication model of the AUV-assisted UVLC data acquisition system includes an AUV, a surface buoy collector, and M USNs, establishing a communication link between the USN and the AUV, and establishing a communication link between the AUV and the surface buoy collector; establishing a channel gain model between the AUV and the USN Establish the channel gain model between AUV and buoy collector According to the navigation time s of the AUV from the starting position to the ending position in the nth time slot n , total time t for AUV to transmit data n , characterizes the total data acquisition delay; by jointly optimizing the multi-dimensional communication resources of spectrum allocation variables, time slot division variables and AUV trajectory planning variables, an AUV-assisted UVLC data acquisition system resource allocation model is established for the optimization problem P1 to minimize the total data acquisition delay; wherein, the time slot division variables include: in the nth time slot, the navigation time s of the AUV from the starting position to the end position n , total time t for AUV to transmit data n , the data collection time β of AUV collecting the mth USN m,n , AUV data upload time γ n ; Spectrum allocation variables include: spectrum allocation ratio α of AUV collected data 1,n And the spectrum allocation ratio α of AUV uploading data 2,n ; AUV trajectory planning variables include: in the nth time slot, the end position q of the AUV's three-dimensional deployment n .

[0007] Preferably, the channel gain model between the AUV and the USN is established as follows: according to the distance d between the AUV and the mth USN in the nth time slot U→A =||u m -q n ||, considering attenuation loss, geometric loss, and small-scale fading that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the m-th USN in the n-th time slot By combining the channel attenuation model between the AUV and the mth USN with the spectrum allocation ratio α of the AUV collected data 1,n , data collection time β m,n , USN transmission power System total bandwidth W and noise power spectral density σ 2 , establish the channel gain model between the AUV and the mth USN in the nth time slot Among them, um Indicates the fixed position of the three-dimensional deployment of the mth USN in the nth time slot; the channel gain model between the AUV and the buoy collector is established Specifically: According to the distance d between the AUV and the buoy collector in the nth time slot A→C =‖q n -v‖, considering attenuation loss, geometric loss, and small-scale fading that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the buoy collector in the nth time slot By combining the channel attenuation model between the AUV and the buoy collector with the spectrum allocation ratio α of the AUV uploaded data 2,n , data upload time γ n , AUV transmission power P A , total system bandwidth W and noise power spectral density σ 2 , establish the channel gain model between AUV and buoy collector in the nth time slot Where v represents the fixed position of the three-dimensional deployment of the buoy collector in the nth time slot.

[0008] Preferably, the resource allocation model of the AUV-assisted UVLC data acquisition system is specifically:

[0009]

[0010] stC1:s n ,t n ,β m,n ,γ n ≥0,0≤α 1,n ,α 2,n ≤1

[0011]

[0012] C3:α 1,n +α 2,n ≤1

[0013]

[0014]

[0015] C8:||q n -q n-1 ‖≤V max s n

[0016] C9:||q n -u m ||≥d min ,||q n -v||≥d min

[0017] In the formula, the optimization target is the total delay of data acquisition n,l∈Τ, time slot set T={1,2,…,N}, N represents the total number of time slots; D m Indicates the amount of data in the mth USN; V max is the maximum speed of AUV; u m , v represent the fixed positions of the three-dimensional deployment of the mth USN and buoy collector in the nth time slot, respectively; d min Indicates the minimum distance to avoid collision.

[0018] According to a second aspect of the present invention, a method for solving a resource allocation model of an AUV-assisted underwater optical communication data acquisition system is provided, and a resource allocation model of an AUV-assisted underwater optical communication data acquisition system constructed according to the modeling method of the resource allocation model of the AUV-assisted underwater optical communication data acquisition system described above is solved.

[0019] Preferably, the solving step includes:

[0020] S6.1. For the optimization problem P1, constraint C7 introduces the Lagrange multiplier λ, constructs the Lagrange function, and transforms the original problem P1 into the dual problem P1-dual.

[0021] S6.2, the optimal value of λ * Introduce the optimization problem P1-dual, so that the optimization problem P1-dual is approximated as the optimization problem P2;

[0022] S6.3, establish n The lower bound s n,min For the optimization problem P1, when the spectrum allocation variables and AUV trajectory planning variables are given, the optimization problem P1 is converted into the time slot partitioning sub-problem P3, and s n,min Substitute the expression into the optimization problem P3 and use the linear programming method to find the optimal solution of the time slot partition variable. The optimal solution is γ n * ;

[0023] S6.4. For the optimization problem P2, when the time slot partitioning variables and the AUV trajectory planning variables are given, the optimization problem P2 is converted into the spectrum allocation sub-problem P4; at the given point that satisfies all the constraints of the spectrum allocation sub-problem P4 On the other hand, the concave function in the concave difference constraint of the spectrum allocation subproblem P4 is and Perform a first-order Taylor expansion and use the linear expression resulting from this Taylor expansion to replace the original concave function in the concave difference constraint of the spectrum allocation subproblem P4, obtaining the optimization problem P5. Based on the optimization problem P5, use the interior point method combined with SCA to solve the spectrum allocation subproblem P4 in polynomial time, obtaining a suboptimal solution to the spectrum allocation subproblem P4.

[0024] S6.5. For the optimization problem P2, when the slot partitioning variables and the spectrum allocation variables are given, the optimization problem P2 is converted into the AUV trajectory planning sub-problem P6; for the AUV trajectory planning sub-problem P6, the concave lower bound is used to approximate the objective function and the constraint C5'. Approximation using concave lower bounds in constraints C4' and C6' And use a convex upper bound to approximate in constraint C4' Approximating with a convex upper bound in constraint C5' For the non-convex constraint C9, a first-order Taylor expansion is used to transform it into a linear constraint to construct the optimization problem P7. Based on the optimization problem P7, the interior point method combined with SCA is used to solve the AUV trajectory planning subproblem P6 in polynomial time to obtain the suboptimal solution of the AUV trajectory planning subproblem P6.

[0025] Preferably, the s n,min The expression is:

[0026] Preferably, an alternating optimization method is used in combination with S6.3-S6.5 to obtain a resource allocation result of the AUV-assisted UVLC data acquisition system.

[0027] According to a third aspect of the present invention, there is provided an AUV-assisted underwater optical communication data acquisition system resource allocation model solving device, comprising a processor and a memory, wherein when the processor executes the computer program stored in the memory, it implements the AUV-assisted underwater optical communication data acquisition system resource allocation model solving method as described in any one of the above.

[0028] According to a fourth aspect of the present invention, a computer-readable storage medium is provided for storing a computer program, wherein when the computer program is executed by a processor, the method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system as described in any one of the above descriptions is implemented.

[0029] The beneficial effects of the present invention are as follows: on the one hand, the present invention constructs a mathematical modeling framework for three-dimensional joint optimization of spectrum-time slot-trajectory by jointly optimizing multi-dimensional communication resources of spectrum allocation variables, time slot division variables and AUV trajectory planning variables, and adopts dynamic optimization of time slot division, spectrum allocation optimization based on the combination of SCA and interior point method, and AUV trajectory planning optimization by using concave-convex double-bound approximation combined with SCA and interior point method, thereby realizing efficient collaborative scheduling of underwater optical communication resources, significantly reducing the total data acquisition delay and greatly improving spectrum utilization. On the other hand, a main problem decomposition and iterative solution strategy is proposed, by converting the optimization problem P1 into P1-dual through dualization and introducing the optimal Lagrange multiplier to convert it into optimization problem P2, and further decomposing P2 into the approximate solution of the spectrum allocation subproblem and the trajectory planning subproblem and the optimal solution of the time slot division subproblem based on p1 for collaborative iteration, thereby effectively reducing the computational complexity of non-convex problems and providing a theoretically rigorous and engineering feasible low-latency optimization solution for underwater mobile optical communication systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A flowchart of the modeling and solution method of the present invention;

[0031] Figure 2 A wireless communication model of an AUV-assisted UVLC data acquisition system provided according to an embodiment of the present invention;

[0032] Figure 3 This is the convergence analysis of the solution method proposed in this invention;

[0033] Figure 4 The time required to complete transmission of each mechanism under different numbers of USNs is shown in Example 2 of the present invention. DETAILED DESCRIPTION

[0034] To make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. It should be noted that, in the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other in any way.

[0035] Example 1: Figure 1-Figure 4As shown, according to a first aspect of an embodiment of the present invention, a modeling method for a resource allocation model of an AUV-assisted underwater optical communication data acquisition system is provided, comprising: constructing a wireless communication model of an AUV-assisted UVLC data acquisition system; the wireless communication model of the AUV-assisted UVLC data acquisition system comprises an AUV, a surface buoy collector and M USNs, a communication link is established between the USN and the AUV, and a communication link is established between the AUV and the surface buoy collector; a channel gain model is established between the AUV and the USN. Establish the channel gain model between AUV and buoy collector According to the navigation time s of the AUV from the starting position to the ending position in the nth time slot n , total time t for AUV to transmit data n , characterizes the total data acquisition delay; by jointly optimizing the multi-dimensional communication resources of spectrum allocation variables, time slot division variables and AUV trajectory planning variables, an AUV-assisted UVLC data acquisition system resource allocation model is established for the optimization problem P1 to minimize the total data acquisition delay; wherein, the time slot division variables include: in the nth time slot, the navigation time s of the AUV from the starting position to the end position n , total time t for AUV to transmit data n , the data collection time β of AUV collecting the mth USN m,n , AUV data upload time γ n ; Spectrum allocation variables include: spectrum allocation ratio α of AUV collected data 1,n and the spectrum allocation ratio α of AUV uploaded data 2,n ; AUV trajectory planning variables include: in the nth time slot, the end position q of the AUV's three-dimensional deployment n .

[0036] Preferably, the channel gain model between the AUV and the USN is established as follows: according to the distance d between the AUV and the mth USN in the nth time slot U→A =||u m -q n ||, considering attenuation loss, geometric loss, and small-scale fading that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the m-th USN in the n-th time slot By combining the channel attenuation model between the AUV and the mth USN with the spectrum allocation ratio α of the AUV collected data 1,n , data collection time β m,n , USN transmission power System total bandwidth W and noise power spectral density σ 2 , establish the channel gain model between the AUV and the mth USN in the nth time slot Among them, um represents the fixed position of the three-dimensional deployment of the m-th USN in the n-th time slot;

[0037] The channel gain model between the AUV and the buoy collector is established Specifically: According to the distance d between the AUV and the buoy collector in the nth time slot A→C =‖q n -v‖, considering attenuation loss, geometric loss, and small-scale fading that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the buoy collector in the nth time slot By combining the channel attenuation model between the AUV and the buoy collector with the spectrum allocation ratio α of the AUV uploaded data 2,n , data upload time γ n , AUV transmission power P A , total system bandwidth W and noise power spectral density σ 2 , establish the channel gain model between AUV and buoy collector in the nth time slot Where v represents the fixed position of the three-dimensional deployment of the buoy collector in the nth time slot.

[0038] Preferably, the resource allocation model of the AUV-assisted UVLC data acquisition system is specifically:

[0039]

[0040] stC1:s n ,t n ,β m,n ,γ n ≥0,0≤α 1,n ,α 2,n ≤1

[0041]

[0042] C3:α 1,n +α 2,n ≤1

[0043]

[0044] C8:‖q n -q n-1 ‖≤V max s n

[0045] C9:||q n -u m ||≥d min , ‖q n -v||≥d min

[0046] In the formula, the optimization target is the total delay of data acquisition n,l∈Τ, time slot set T={1,2,…,N}, N represents the total number of time slots; D m Indicates the amount of data in the mth USN; V max is the maximum speed of AUV; u m , v represent the fixed positions of the three-dimensional deployment of the mth USN and buoy collector in the nth time slot, respectively; d min Indicates the minimum distance to avoid collision.

[0047] According to the second aspect of an embodiment of the present invention, a method for solving a resource allocation model of an AUV-assisted underwater optical communication data acquisition system is provided, and the resource allocation model of an AUV-assisted underwater optical communication data acquisition system constructed according to the modeling method of the resource allocation model of the AUV-assisted underwater optical communication data acquisition system described above is solved.

[0048] Preferably, the solving step includes:

[0049] S6.1. For the optimization problem P1, constraint C7 introduces the Lagrange multiplier λ, constructs the Lagrange function, and transforms the original problem P1 into the dual problem P1-dual.

[0050] S6.2, the optimal value of λ * Introduce the optimization problem P1-dual, so that the optimization problem P1-dual is approximated as the optimization problem P2;

[0051] S6.3, establish n The lower bound s n,min For the optimization problem P1, when the spectrum allocation variables and AUV trajectory planning variables are given, the optimization problem P1 is converted into the time slot partitioning sub-problem P3, and s n,min Substitute the expression into the optimization problem P3 and use the linear programming method to find the optimal solution of the time slot partition variable. The optimal solution is γ n * ;

[0052] S6.4. For the optimization problem P2, when the time slot partitioning variables and the AUV trajectory planning variables are given, the optimization problem P2 is converted into the spectrum allocation sub-problem P4; at the given point that satisfies all the constraints of the spectrum allocation sub-problem P4 On the other hand, the concave function in the concave difference constraint of the spectrum allocation subproblem P4 is and Perform a first-order Taylor expansion and use the linear expression resulting from this Taylor expansion to replace the original concave function in the concave difference constraint of the spectrum allocation subproblem P4, obtaining the optimization problem P5. Based on the optimization problem P5, use the interior point method combined with SCA to solve the spectrum allocation subproblem P4 in polynomial time, obtaining a suboptimal solution to the spectrum allocation subproblem P4.

[0053] S6.5. For the optimization problem P2, when the slot partitioning variables and the spectrum allocation variables are given, the optimization problem P2 is converted into the AUV trajectory planning sub-problem P6; for the AUV trajectory planning sub-problem P6, the concave lower bound is used to approximate the objective function and the constraint C5'. Approximation using concave lower bounds in constraints C4' and C6' And use a convex upper bound to approximate in constraint C4' Approximating with a convex upper bound in constraint C5' For the non-convex constraint C9, a first-order Taylor expansion is used to transform it into a linear constraint to construct the optimization problem P7. Based on the optimization problem P7, the interior point method combined with SCA is used to solve the AUV trajectory planning subproblem P6 in polynomial time to obtain the suboptimal solution of the AUV trajectory planning subproblem P6.

[0054] Preferably, the s n,min The expression is:

[0055] Preferably, an alternating optimization method is used in combination with S6.3-S6.5 to obtain a resource allocation result of the AUV-assisted UVLC data acquisition system.

[0056] According to a third aspect of an embodiment of the present invention, a device for solving a resource allocation model of an AUV-assisted underwater optical communication data acquisition system is provided, comprising a processor and a memory, wherein when the processor executes a computer program stored in the memory, a method for solving a resource allocation model of an AUV-assisted underwater optical communication data acquisition system as described in any one of the above descriptions is implemented.

[0057] According to a fourth aspect of an embodiment of the present invention, a computer-readable storage medium is provided for storing a computer program, wherein when the computer program is executed by a processor, the method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system as described in any one of the above descriptions is implemented.

[0058] Example 2: The following is an optional specific implementation of the present invention with reference to the accompanying drawings:

[0059] like Figure 1-Figure 4 As shown, a resource allocation method in an AUV-assisted UVLC data acquisition system includes the following steps:

[0060] S1, build the wireless communication model of the AUV-assisted UVLC data acquisition system and initialize the system parameters;

[0061] The wireless communication model of the AUV-assisted UVLC data acquisition system includes an AUV, a surface buoy collector and M USNs. A communication link is established between the USN and the AUV, and a communication link is established between the AUV and the surface buoy collector. Each USN and AUV are deployed with the same transmitter, and the AUV and the buoy collector are deployed with the same receiver. The USN transmitter sends data to the AUV receiver through a water optical channel, and the AUV transmitter sends data to the buoy collector receiver through a water optical channel.

[0062] Specifically: For the wireless communication model of the AUV-assisted UVLC data acquisition system, both the underwater access link and the underwater backhaul link adopt frequency division multiple access technology for wireless communication; the fixed positions of the three-dimensional deployment of the mth USN and the buoy collector are denoted as u m and v; define the USN set as I = {1, 2, ..., M} (m∈I); the complete data transmission process is divided into time slots, and the time slot set is T = {1, 2, ..., N} (n, l∈Τ); in the nth time slot, the end position of the AUV three-dimensional deployment is q n .

[0063] Exemplarily, the initialization system parameters include the USN initial position coordinates, the AUV initial position coordinates, the position coordinates of the buoy collector, the maximum navigation speed of the AUV, the cache capacity of the AUV, the data storage capacity of all USNs, the minimum safety distance between the AUV and the USN, the minimum safety distance between the AUV and the buoy collector, etc.

[0064] S2, construct the channel gain model between AUV and USN;

[0065] The channel gain model between the AUV and the USN is constructed, including: according to the distance d between the AUV and the mth USN in the nth time slot U→A =||u m -q n ||, considering attenuation loss Geometric loss The small-scale fading h obeys the log-normal distribution and constructs the channel attenuation model between the AUV and the m-th USN in the n-th time slot By combining the channel attenuation model between the AUV and the mth USN with the spectrum allocation ratio α of the AUV collected data 1,n , data collection time β m,n , USN transmission power System total bandwidth W and noise power spectral density σ 2, establish the channel gain model between AUV and USN; where e represents the natural constant, c is the extinction coefficient, D R is the receiver aperture diameter, θ 1 / e is the beam divergence angle.

[0066] The channel attenuation model between the AUV and the mth USN is combined with the spectrum allocation ratio α of the AUV collected data. 1,n , data collection time β m,n , USN transmission power System total bandwidth W and noise power spectral density σ 2 , establish the channel gain model between AUV and USN, the expression is:

[0067]

[0068] in, is the channel gain model between the AUV and the mth USN in the nth time slot, and the numerator is Represents the received signal power, the denominator a 1,n Wσ 2 is the equivalent noise power, and the two constitute the signal-to-noise ratio. Finally, the channel gain model between AUV and USN is derived through the Shannon capacity formula.

[0069] The small-scale fading that obeys the log-normal distribution is obtained according to the probability density function f(h), and the probability density function f(h) is expressed as:

[0070]

[0071] Among them, the mean variance is the flicker index.

[0072] The scintillation index The expression is:

[0073]

[0074] in, is the wave number, λ is the wavelength of light, κ is the spatial frequency, and ζ is the normalized path length integral variable; Φ(κ) is the spatial power spectrum of the ocean optical turbulence refractive index, which is related to the temperature and salinity of the water. When the eddy heat diffusivity and the eddy salt diffusivity are equal, its mathematical expression is:

[0075]

[0076] Where ε is the turbulent kinetic energy dissipation rate per unit mass of fluid, α is the Kolmogorov microscale, χT is the mean square temperature dissipation rate, and ω is the relative strength of temperature fluctuation and salinity fluctuation; A T 、A S and A TS are all model parameters, all positive numbers;

[0077] S3, building a model of the channel gain between the AUV and the buoy collector;

[0078] The channel gain model between the AUV and the buoy collector is constructed, including: according to the distance d between the AUV and the buoy collector in the nth time slot A→c =||q n -v||, considering attenuation loss, geometric loss, and small-scale fading h that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the buoy collector in the nth time slot By combining the channel attenuation model between the AUV and the buoy collector with the spectrum allocation ratio α of the AUV uploaded data 2,n , data upload time γ n , AUV transmission power P A , total system bandwidth W and noise power spectral density σ 2 , establish the channel gain model between AUV and buoy collector.

[0079] It should be noted that the channel attenuation model between the AUV and the buoy collector is The establishment method is the same as the channel attenuation model between the AUV and the mth USN.

[0080] Furthermore, the channel attenuation model between the AUV and the buoy collector is combined with the spectrum allocation ratio α of the AUV uploaded data. 2,n , data upload time γ n , AUV transmission power P A , total system bandwidth W and noise power spectral density σ 2 , establish the channel gain model between AUV and buoy collector, the expression is:

[0081]

[0082] in, is the channel gain model between the AUV and the buoy collector in the nth time slot, and the numerator is Represents the received signal power, the denominator a 2,n Wσ 2 is the equivalent noise power, and the two constitute the signal-to-noise ratio. Finally, the channel gain model between the AUV and the buoy collector is derived through the Shannon capacity formula.

[0083] S4, characterizes the total delay of data acquisition;

[0084] The total delay of data acquisition Among them, s n , t n They are the navigation time of the AUV from the starting position to the ending position and the total time of the AUV transmitting data in the nth time slot, respectively.

[0085] S5, by jointly optimizing the multi-dimensional communication resources of spectrum allocation variables, time slot division variables and AUV trajectory planning variables, a resource allocation model of the AUV-assisted UVLC data acquisition system is established for the optimization problem P1 to minimize the total data acquisition delay;

[0086] The construction of the resource allocation model of the AUV-assisted UVLC data acquisition system is specifically as follows: in the construction of the AUV-assisted UVLC data acquisition system, by jointly optimizing the multi-dimensional communication resources of the spectrum allocation variable, the time slot division variable and the AUV trajectory planning variable, the AUV-assisted UVLC data acquisition system resource allocation model is established for the optimization problem P1 to minimize the total data acquisition delay; in the above, the time slot division variable involves: in the nth time slot, the navigation time s of the AUV from the starting position to the ending position n , total time t for AUV to transmit data n , the data collection time β of AUV collecting the mth USN m,n , AUV data upload time γ n ; The spectrum allocation variable involves the spectrum allocation ratio α of the AUV collected data 1,n and the spectrum allocation ratio α of AUV uploaded data 2,n ; AUV trajectory planning variables involve: in the nth time slot, the terminal position q of the AUV's three-dimensional deployment n .

[0087] The resource allocation model of the AUV-assisted UVLC data acquisition system is expressed as:

[0088]

[0089] stC1:s n ,t n ,β m,n ,γ n ≥0,0≤α 1,n ,α 2,minn ≤1

[0090]

[0091] C3:α 1,n +α 2,n ≤1

[0092]

[0093] C8:||q n -q n-1 ||≤V max s n

[0094] C9:||q n -u m ||≥d min ,||q n -v||≥d min

[0095] In the formula, the goal of the optimization problem P1 is to minimize the total delay of data acquisition, and the optimization goal is q n Indicates the final position of the AUV 3D deployment in the nth time slot; u m , v represent the fixed positions of the three-dimensional deployment of the mth USN and buoy collector in the nth time slot respectively; constraint C1 requires that all time-related variables are non-negative, that is, s n , t n , β m,n , γ n are all non-negative values; the spectrum allocation ratio α required for AUV to collect data 1,n and the spectrum allocation ratio α for upload data 2,n are all between 0 and 1; constraint C2 requires the total time t for AUV to transmit data n The time when the AUV collects all USN data in the nth time slot and data upload time γ n The larger value; Constraint C3 requires the spectrum allocation ratio α of the collected data 1,n and the spectrum allocation ratio α for upload data 2,n The sum is less than or equal to 1; Constraint C4 requires that the amount of data collected by the AUV on the mth USN from the starting time slot to any time slot l More than the total uploaded data Constraint C5 requires that the amount of data collected by the AUV on the mth USN from the starting time slot to any time slot l is The cumulative amount of uploaded data The difference cannot exceed the AUV cache capacity B; Constraint C6 requires that the amount of data collected by the AUV for the mth USN in the entire time slot Must be greater than or equal to the amount of data D in the mth USN m , that is, the AUV must complete the data collection in any USN; constraint C7 requires the total amount of data uploaded by the AUV to the buoy collector in the entire time slot Must be greater than or equal to the total amount of data in all USNs That is, the AUV must collect all the data in the USN to avoid data loss due to buffer overflow or transmission interruption; constraint C8 requires the AUV to move a distance of ||q in the time slot n -q n-1 || must be less than or equal to the maximum distance the AUV can move, which is the navigation time s n With the maximum speed of AUV V max Constraint C9 requires that the distance between the AUV and the m-th USN in the n-th time slot || q n -u m ||, the distance between the AUV and the buoy collector in the nth time slot ||q n -v|| must be greater than or equal to the minimum distance d to prevent collision min .

[0096] By applying the above technical solution, it can be seen that the resource allocation model of the AUV-assisted UVLC data acquisition system is based on the three-dimensional trajectory planning variable q n and spectrum allocation variable α 1,n , α 2,n The collaborative optimization breaks through the limitations of traditional static resource allocation and introduces the time slot partition variable s n , t n , β m,n , γ n To achieve dimensional control of communication time and navigation time, constraints C4-C9 systematically establish a quantitative balance relationship between USN data acquisition, AUV cache dynamics and buoy data return, providing a theoretically complete and engineering feasible modeling method for end-to-end delay optimization of underwater mobile communication systems.

[0097] S6, solving the resource allocation model of the AUV-assisted UVLC data acquisition system;

[0098] Said S6 is specifically:

[0099] S6.1. For the optimization problem P1, constraint C7 introduces the Lagrange multiplier λ (λ ≥ 0), constructs the Lagrange function, and transforms the original problem P1 into the dual problem P1-dual.

[0100] The mathematical model of the optimization problem P1-dual is:

[0101]

[0102] Θ={s n ,t n ,β m,n ,γ n ,α 1,n ,α 2,n,q n}

[0103] stC1:s n ,t n ,β m,n ,γ n ≥0,0≤α 1,n ,α 2,n ≤1

[0104]

[0105] C3:α 1,n +α 2,n ≤1

[0106]

[0107] C8:||q n -q n-1 ||≤V max s n

[0108] C9:||q n -u m ||≥d min ,||q n -v||≥d min

[0109] S6.2, the optimal value of λ * Introduce the optimization problem P1-dual, so that the optimization problem P1-dual is approximated as the optimization problem P2;

[0110] The objective function of the optimization problem P2 is: The constraints of P2 are the same as those of P1-dual. The optimization problem P2 is still a non-convex problem.

[0111] In the following, the time slot partitioning subproblem will be constructed based on P1, and the spectrum allocation subproblem and AUV trajectory planning subproblem will be constructed based on P2.

[0112] S6.3, establish n The lower bound s n,min For the optimization problem P1, when the spectrum allocation variables and AUV trajectory planning variables are given, the optimization problem P1 is converted into the time slot partitioning sub-problem P3, and s n,min Substitute the expression into the optimization problem P3 and use the linear programming method to find the optimal solution of the time slot partition variable. The optimal solution is γ n * ; Among them, s n,min Indicates s nThe lower bound, q n-1 The final position of the AUV's 3D deployment in the n-1th time slot.

[0113] The mathematical model of the sub-problem P3 is:

[0114]

[0115] in:

[0116]

[0117] Since the optimization variable s n Monotonically increasing, so It must be s n The lower bound of s n,min Substituting into the optimization problem P3, the optimization problem P3 becomes a linear programming problem, from which we can further use the mature linear programming method to obtain its optimal solution in polynomial time. The optimal solution is recorded as γ n * .

[0118] S6.4. For the optimization problem P2, when the time slot partitioning variables and the AUV trajectory planning variables are given, the optimization problem P2 is converted into the spectrum allocation sub-problem P4; at the given point that satisfies all the constraints of the spectrum allocation sub-problem P4 On the other hand, the concave function in the concave difference constraint of the spectrum allocation subproblem P4 is and Perform a first-order Taylor expansion and use the linear expression after the Taylor expansion to replace the original concave function in the concave difference constraint of the spectrum allocation subproblem P4, and obtain the convex optimization problem P5. Based on the convex optimization problem P5, the spectrum allocation subproblem P4 is solved in polynomial time using the interior point method combined with SCA to obtain the suboptimal solution of the spectrum allocation subproblem P4. Indicates that the time slot division variable and the AUV trajectory planning variable are given, with α 1,n As a variable Indicates that the time slot division variable and the AUV trajectory planning variable are given, with α 2,n As a variable

[0119] The mathematical model of the sub-problem P4 is:

[0120]

[0121] stC1 * :0≤α 1,n ,α 2,n ≤1

[0122] C3:α 1,n +α 2,n ≤1

[0123]

[0124] in:

[0125]

[0126] further, for The second derivative of for The second-order derivative of is expressed as follows:

[0127]

[0128] The second-order derivatives show that both are concave functions, due to the constraint C4 * With C5 * Since it involves the difference of concave functions, we know that the constraint C4 in P4 is * With C5 * It belongs to concave difference programming; in order to deal with the non-convexity of the constraint function, the Successive Convex Approximation (SCA) algorithm is used to solve the problem of the given point satisfying all the constraints of P4. On, will and Perform a first-order Taylor expansion and use the linear expression after the Taylor expansion in C4 * With C5 * Substituting the original function in, we get the convex optimization problem P5; and The first-order Taylor expansion expression of is:

[0129]

[0130] in:

[0131]

[0132] Where, They are At a given point The first derivative at a given point All constraints of P4 are met.

[0133] This method can be used to approximate the optimization problem P4 to the convex optimization problem P5. The expression of the convex optimization problem P5 is:

[0134]

[0135] stC1 * :0≤α 1,n ,α 2,n ≤1

[0136] C3:α 1,n +α 2,n ≤1

[0137]

[0138] Based on the convex optimization problem P5, the spectrum allocation subproblem P4 is solved in polynomial time using the interior point method combined with SCA to obtain the suboptimal solution of the spectrum allocation subproblem P4. The specific solution process is as follows:

[0139] Step 6.4.1, Initialization: Set the initial iteration number i to 1, set the allowable error value ∈1 to a small positive number; set the initial given point that satisfies all constraints of the optimization problem P4 Set as the initial feasible solution and set the initial objective function value F (1) for That is, α in the first iteration 1,n , α 2,n .

[0140] Step 6.4.2: At a given point In the case of , the interior point method is used to solve the optimization problem P5, and the optimal solution is recorded as Will Substitute the objective function of P4 and get F * .

[0141] Step 6.4.3: If F * -F (i) <∈1, the algorithm converges, is the approximate solution sought; otherwise, add 1 to the number of iterations i and update F (i) =F * , return to Step 6.4.2.

[0142] Therefore, the present invention gradually approaches the optimal solution of the non-convex optimization problem P4 by continuously iterating the convex optimization problem P5, and obtains the suboptimal solution of the optimization problem P4, which is recorded as

[0143] S6.5. For the optimization problem P2, when the slot partitioning variables and the spectrum allocation variables are given, the optimization problem P2 is converted into the AUV trajectory planning sub-problem P6; for the AUV trajectory planning sub-problem P6, the concave lower bound is used to approximate the objective function and the constraint C5'. Approximation using concave lower bounds in constraints C4' and C6' And use a convex upper bound to approximate in constraint C4' Approximating with a convex upper bound in constraint C5' For the non-convex constraint C9, a first-order Taylor expansion is used to transform it into a linear constraint to construct the optimization problem P7. Based on the optimization problem P7, the interior point method combined with SCA is used to solve the AUV trajectory planning sub-problem P6 in polynomial time to obtain the suboptimal solution of the AUV trajectory planning sub-problem P6. Among them, the constraints C4', C5', and C6' are the constraints C4, C5, and C6 in the optimization problem P2 according to the "time slot division variable and spectrum allocation variable given, with q n The corresponding constraints after the "variable" condition changes; among them, Indicates that the time slot division variable and spectrum allocation variable are given, with q n As a variable Indicates that the time slot division variable and spectrum allocation variable are given, with q n As a variable

[0144] The mathematical model of the sub-problem P6 is:

[0145]

[0146] C8:||q n -q n-1 ||≤V max s n

[0147] C9:||q n -u m ||≥d min ,||q n -v||≥d min

[0148] in:

[0149]

[0150] according to The second-order Hessian matrix characteristics of the two are non-concave and non-convex, which leads to the AUV trajectory planning sub-problem P6 is still a non-convex optimization problem. In order to transform the optimization problem P6 into a convex problem while ensuring a feasible solution, a concave lower bound is used to approximate the objective function and constraint C5'. In constraints C4' and C6', use The concave lower bound of the approximation is used, and the convex upper bound is used to approximate the constraint C4'. Approximating with a convex upper bound in constraint C5' By controlling the approximation error by combining the upper and lower bounds, the feasible solution can be strictly guaranteed to be within the original problem. The corresponding upper and lower bounds; the present invention uses SCA to solve the optimization problem P6;.

[0151] First, in order to find The concave lower bound of n =||q n -v|| 2 ,definition

[0152] The first-order derivative and second-order derivative expressions are:

[0153] It can be seen that About d n Monotonically decreasing and strictly concave, at a given point that satisfies all constraints of the AUV trajectory planning subproblem P6 Up (given point satisfy right Perform a first-order Taylor expansion to find the concave lower bound, The first-order Taylor expansion of is:

[0154]

[0155] d n =||q n -v|| 2 Substituting into the first-order Taylor expansion, we get At a given point on the AUV The concave lower bound equivalent of :

[0156]

[0157] Due to d n =||q n -v|| 2 is a convex function, and is a monotonically decreasing function, so is a concave function.

[0158] Furthermore, by the same method, at a given point right Performing a first-order Taylor expansion, we can obtain The concave lower bound equivalent of is expressed as:

[0159]

[0160] Where, d m,n =||q n -u m ||2 .

[0161] Second, in order to find and The convex upper bound of is obtained. Based on its second-order Taylor expansion, a positive definite matrix is used to approximate the global upper bound of the non-positive definite Hessian matrix. The second-order Taylor expansion expression of is:

[0162]

[0163] in, and They are At a given point The first-order derivative and second-order derivative matrix at , calculated using the chain rule and The expanded form of is:

[0164]

[0165]

[0166] Where, and They are d n (q n )=||q n -v|| 2 The first-order derivative and second-order derivative matrix of , I is a 3×3 size identity matrix; in In the expansion of The only non-zero eigenvalues of Further, due to The second matrix in the expansion of is a diagonal matrix. According to the properties of the diagonal matrix, we can The eigenvalues of are equivalent to Further, combined and The specific expansion of can be obtained The eigenvalue vector of , the eigenvector expression is:

[0167]

[0168] in,

[0169] exist In the eigenvalue vector V of , expand the first eigenvalue, the expression is:

[0170]

[0171] It can be seen that the first eigenvalue in V is positive, according to It can be seen that the rest of the eigenvalues in V are negative;

[0172] Furthermore, the Hessian matrix is diagonalized by the orthogonal matrix Q, and the expression is:

[0173]

[0174] Where Q is The eigenvector concatenation of the first term in diag(V) is The diagonal matrix of the eigenvalue vectors V.

[0175] Furthermore, we define the positive definite matrix is the global positive upper bound of the Hessian matrix. Therefore, The second-order term of the Taylor expansion of Through this transformation, we obtain The convex upper bound equivalent of is expressed as:

[0176]

[0177] Further, using the same method, we can easily get The convex upper bound of is expressed as:

[0178]

[0179] Where, in

[0180] So through the above steps, we get The convex upper bound of .

[0181] Third, for the remaining non-convex constraint C9, a tighter linear constraint is obtained by performing a first-order Taylor expansion on the convex constraint function. The specific expression is:

[0182]

[0183] Through the above steps, the AUV trajectory planning sub-problem P6 can be transformed into a convex optimization problem P7, and the expression of P7 is:

[0184] C8:||q n -q n-1 ||≤V max s n

[0185]

[0186] For the convex optimization problem P7, the interior point method can be used to obtain a suboptimal solution in polynomial time. Furthermore, in order to improve the optimization performance, the obtained solution can be replaced by Then continue to solve the updated optimization problem P7; based on the SCA algorithm and through multiple iterations, convergence obtains a suboptimal solution that meets the Karush-Kuhn-Tucker optimality conditions of the optimization problem P6; since the original feasible domain has been replaced by a more conservative feasible domain, the solution of each iteration must satisfy the original constraints of the AUV trajectory planning subproblem P6; the process of solving the AUV trajectory planning subproblem is as follows:

[0187] Step 6.5.1, Initialization: Set the initial iteration number j to 1, set the allowable error value ∈2 to a small positive number, and set the initial given point that satisfies all constraints of the AUV trajectory planning subproblem P6 Assume that the feasible solution of the AUV trajectory planning subproblem P6, and the initial target value F (1) Set to Then proceed to Step 6.5.2.

[0188] Step 6.5.2: Given Under these conditions, the optimization problem P7 is solved based on the interior point method, and the optimal solution is recorded as Will Substituted into the objective function of P6, we get

[0189] Step 6.5.3: If F * -F (j) <∈2, the algorithm converges, This is the approximate solution to the optimization problem P6; otherwise, add 1 to the value of the iteration number j and update F (j) =F * , return to Step 6.5.2.

[0190] It should be noted that after the transformation of the constraint C4 under the optimization problem P1 / P2, the subsequent optimization problems P3, P4, P5, P6, P7 and C4 * 、C4 # , C4', and C4"; for other constraints under the optimization problem P1-dual / P2 that have transformations, they are distinguished in a similar way.

[0191] S6.6. Combine S6.3-S6.5 above and use the alternating optimization method to obtain the resource allocation result of the AUV-assisted UVLC data acquisition system.

[0192] Furthermore, the S6.6 is specifically as follows:

[0193] S6.6.1: Initialization: Set k = 1 to the current number of iterations, ∈3 to the allowable error value, and set the initial target value G (0) +∞; given the initial variables that satisfy the constraints of the time slot partitioning subproblem P3 Perform S6.6.2;

[0194] S6.6.2: In the current optimization variable Under these conditions, the optimal solution of the time slot partitioning subproblem P3 is obtained by the linear programming method described in S6.3 and updated to Perform S6.6.3;

[0195] S6.6.3: In the current optimization variable Under these conditions, the suboptimal solution of spectrum allocation subproblem P4 is obtained by combining the interior point method described in S6.4 with the SCA method and updated to Implement S6.6.4;

[0196] S6.6.4: In the current optimization variable Under these conditions, the suboptimal solution of the AUV trajectory planning subproblem P6 is obtained by combining the interior point method described in S6.5 with the SCA method and updated to Implement S6.6.5;

[0197] S6.6.5: If G (k-1) -G (k) ≤∈3, the proposed algorithm converges, and the current This is the suboptimal solution to the optimization problem P1; otherwise, set k = k + 1 and return to S6.6.2.

[0198] Furthermore, combined with experimental data, the following is explained: the present invention considers a circular area with a depth of 0-100m and a radius of 50m; the USN is distributed at a fixed depth of -100m and the position is random in the circular area; the buoy collector is located at the center of the water surface area; the simulation experiment is carried out according to the specific experimental steps of the present invention, and the parameters involved are shown in Table 1.

[0199] Table 1 Parameter setting table

[0200]

[0201] Figure 3The convergence analysis of the algorithm of the present invention for 10 USN nodes in 100 experiments was demonstrated. In an optimization problem with the objective function of minimizing the total data acquisition delay, it can be observed that the total delay gradually decreases and eventually stabilizes with increasing iterations, verifying the convergence of the algorithm. The convergence of the algorithm is directly related to the optimization effect of minimum data delay. The fast convergence of the algorithm shows that the present invention can approach the theoretical optimal solution with fewer iterations, significantly reducing computational time. The stable convergence characteristics ensure the accuracy of the resource scheduling strategy, avoid delay jitter caused by oscillation or local optimality, and ultimately achieve the optimization goal of global low latency.

[0202] Figure 4 The results show that the present invention (the obtained mechanism) is compared with the following two comparison mechanisms under 100 random network topologies: Comparison mechanism 1 adopts fixed spectrum allocation and sets the optimization variable α of spectrum allocation to 1,n Randomly select values between 0.2 and 0.8, α 2,n The value is 1-α 1,n , the rest of the parts are consistent with the algorithm proposed in this invention; the comparison mechanism 2 uses a fixed AUV track, which is a uniform polygonal motion along a circular path with a radius of 45m at an altitude of -50m, and the rest of the parts are consistent with the algorithm proposed in this invention. Figure 4 It can be seen that compared with comparison mechanism 1, the algorithm of the present invention significantly reduces the total delay by dynamically optimizing spectrum allocation. When the number of USNs increases from 8 to 16, the total delay of comparison mechanism 1 shows a significant upward trend due to the inability of random spectrum allocation to adapt to the heterogeneous requirements of the equipment; when the number of USNs reaches 16, the algorithm of the present invention is reduced by an average of 119.07 seconds compared with comparison mechanism 1, confirming the effectiveness of dynamic resource allocation in high-density scenarios. Compared with comparison mechanism 2, the dynamic trajectory planning of the mechanism proposed in the present invention further highlights the performance advantage. Specifically, the average total delay of the proposed mechanism is reduced by 197.04 seconds compared with comparison mechanism 2. This result fully demonstrates that the proposed algorithm can effectively cope with the challenges of dense access of large-scale USN devices by jointly optimizing spectrum allocation, time slot division and AUV trajectory planning, and achieve continuous improvement of system performance.

[0203] The specific embodiments of the present invention are described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Various changes can be made within the knowledge of ordinary technicians in this field without departing from the scope of the present invention.

Claims

1. A modeling method for a resource allocation model of an AUV-assisted underwater optical communication data acquisition system, characterized in that: include: Build a wireless communication model for the AUV-assisted UVLC data acquisition system; The wireless communication model of the AUV-assisted UVLC data acquisition system includes an AUV, a surface buoy collector and M USNs, a communication link is established between the USN and the AUV, and a communication link is established between the AUV and the surface buoy collector; Establishing the channel gain model between AUV and USN Establish the channel gain model between AUV and buoy collector According to the navigation time s of the AUV from the starting position to the ending position in the nth time slot n , total time t for AUV to transmit data n , characterizes the total delay of data acquisition; By jointly optimizing the multi-dimensional communication resources of spectrum allocation variables, time slot division variables and AUV trajectory planning variables, a resource allocation model of the AUV-assisted UVLC data acquisition system is established for the optimization problem P1 to minimize the total data acquisition delay; wherein the time slot division variables include: the navigation time s of the AUV from the starting position to the ending position in the nth time slot n , total time t for AUV to transmit data n , the data collection time β of AUV collecting the mth USN m,n , AUV data upload time γ n ; Spectrum allocation variables include: spectrum allocation ratio α of AUV collected data 1,n And the spectrum allocation ratio α of AUV uploading data 2,n ; AUV trajectory planning variables include: in the nth time slot, the end position q of the AUV's three-dimensional deployment n .

2. The modeling method of the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to claim 1 is characterized in that: The channel gain model between the AUV and the USN is established as follows: According to the distance d between the AUV and the mth USN in the nth time slot U→A =||u m -q n ||, considering attenuation loss, geometric loss, and small-scale fading that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the m-th USN in the n-th time slot By combining the channel attenuation model between the AUV and the mth USN with the spectrum allocation ratio α of the AUV collected data 1,n , data collection time β m,n , USN transmission power System total bandwidth W and noise power spectral density σ 2 , establish the channel gain model between the AUV and the mth USN in the nth time slot Among them, u m represents the fixed position of the three-dimensional deployment of the m-th USN in the n-th time slot; The channel gain model between the AUV and the buoy collector is established Specifically: According to the distance d between the AUV and the buoy collector in the nth time slot A→C =||q n -v||, considering attenuation loss, geometric loss, and small-scale fading that obeys the log-normal distribution, a channel attenuation model is constructed between the AUV and the buoy collector in the nth time slot By combining the channel attenuation model between the AUV and the buoy collector with the spectrum allocation ratio α of the AUV uploaded data 2,n , data upload time γ n , AUV transmission power P A , total system bandwidth W and noise power spectral density σ 2 , establish the channel gain model between AUV and buoy collector in the nth time slot Where v represents the fixed position of the three-dimensional deployment of the buoy collector in the nth time slot.

3. The modeling method of the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to claim 1 is characterized in that: The resource allocation model of the AUV-assisted UVLC data acquisition system is specifically as follows: stC1:s n ,t n ,b m,n ,c n ≥0.0≤α 1,n ,a 2,n ≤1 C3:α 1,n +α 2,n ≤1 C8:||q n -q n-1 ||≤V max s n C9:||q n -u m ||≥d min ,||q n -v||≥d min In the formula, the optimization target is the total delay of data acquisition Time slot set T = {1, 2, ..., N}, where N represents the total number of time slots; D m Indicates the amount of data in the mth USN; V max is the maximum speed of AUV; u m , v represent the fixed positions of the three-dimensional deployment of the mth USN and buoy collector in the nth time slot, respectively; d min Indicates the minimum distance to avoid collision.

4. A method for solving a resource allocation model for an AUV-assisted underwater optical communication data acquisition system, characterized in that: The AUV-assisted underwater optical communication data acquisition system resource allocation model constructed according to the modeling method of the AUV-assisted underwater optical communication data acquisition system resource allocation model according to claim 3 is solved.

5. The method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to claim 4 is characterized in that: The solving step comprises: S6.

1. For the optimization problem P1, constraint C7 introduces the Lagrange multiplier λ, constructs the Lagrange function, and transforms the original problem P1 into the dual problem P1-dual. S6.2, the optimal value of λ * Introduce the optimization problem P1-dual, so that the optimization problem P1-dual is approximated as the optimization problem P2; S6.3, establish n The lower bound s n,min For the optimization problem P1, when the spectrum allocation variables and AUV trajectory planning variables are given, the optimization problem P1 is converted into the time slot partitioning sub-problem P3, and s n,min Substitute the expression into the optimization problem P3 and use the linear programming method to find the optimal solution of the time slot partition variable. The optimal solution is γ n * ; S6.

4. For the optimization problem P2, when the time slot partitioning variables and the AUV trajectory planning variables are given, the optimization problem P2 is converted into the spectrum allocation sub-problem P4; at the given point that satisfies all the constraints of the spectrum allocation sub-problem P4 On the other hand, the concave function in the concave difference constraint of the spectrum allocation subproblem P4 is and Perform a first-order Taylor expansion and use the linear expression resulting from this Taylor expansion to replace the original concave function in the concave difference constraint of the spectrum allocation subproblem P4, obtaining the optimization problem P5. Based on the optimization problem P5, use the interior point method combined with SCA to solve the spectrum allocation subproblem P4 in polynomial time, obtaining a suboptimal solution to the spectrum allocation subproblem P4. S6.

5. For the optimization problem P2, when the slot partitioning variables and the spectrum allocation variables are given, the optimization problem P2 is converted into the AUV trajectory planning sub-problem P6; for the AUV trajectory planning sub-problem P6, the concave lower bound is used to approximate the objective function and the constraint C5'. Approximation using concave lower bounds in constraints C4' and C6' And use a convex upper bound to approximate in constraint C4' Approximating with a convex upper bound in constraint C5' For the non-convex constraint C9, a first-order Taylor expansion is used to transform it into a linear constraint to construct the optimization problem P7. Based on the optimization problem P7, the interior point method combined with SCA is used to solve the AUV trajectory planning subproblem P6 in polynomial time to obtain the suboptimal solution of the AUV trajectory planning subproblem P6.

6. The method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to claim 5 is characterized in that: The s n,min The expression is:

7. The method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to claim 5, characterized in that: Combined with S6.3-S6.5, the alternating optimization method is used to solve the resource allocation result of the AUV-assisted UVLC data acquisition system.

8. A device for solving resource allocation models of AUV-assisted underwater optical communication data acquisition systems, characterized in that: The method comprises a processor and a memory, wherein when the processor executes the computer program stored in the memory, it implements the method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to any one of claims 4 to 7.

9. A computer-readable storage medium, characterized in that Used to store a computer program, wherein when the computer program is executed by a processor, it implements the method for solving the resource allocation model of the AUV-assisted underwater optical communication data acquisition system according to any one of claims 4 to 7.

Images