Multi-domain resource scheduling method and device for RIS-assisted satellite Internet of Things (ISAC)

Abstract

The invention relates to the technical field of wireless communication, and provides an RIS-assisted satellite internet of things ISAC multi-domain resource scheduling method and device. An optimization problem is decomposed through an alternative optimization framework, different solving strategies are adopted for different RIS types, and unified equivalent channel modeling is carried out, so that the constructed RIS-assisted satellite Internet of Things ISAC system can be compatible with traditional diagonal RIS and super diagonal RIS, a unified system model is provided to give full play to different RIS architectures, and the construction efficiency of the RIS-assisted satellite Internet of Things ISAC system is improved. And the complex and changeable communication and perception requirements in a satellite Internet of Things scene are met. A super-diagonal RIS phase shift optimization sub-problem is solved by combining Riemannian gradient descent with a Givens rotation search strategy, so that a non-convex constraint problem of an Euclidean space is converted into an unconstrained optimization problem on a Riemannian manifold, and the solving efficiency and the optimization precision are remarkably improved.

Description

Technical Field

[0001] This application relates to the field of wireless communication technology, and in particular to a RIS-assisted satellite Internet of Things (ISAC) multi-domain resource scheduling method and apparatus. Background Technology

[0002] With the rapid development of sixth-generation (6G) wireless communication technology, Integrated Sensing and Communication (ISAC) has become a key enabling technology for future wireless networks. ISAC technology achieves deep integration of communication and sensing functions by sharing spectrum resources and hardware devices, significantly improving spectrum efficiency and system performance.

[0003] Low Earth Orbit (LEO) satellite communication systems, with their advantages of low latency and wide coverage, have become an important component in building an integrated space-air-ground network. Satellite Internet of Things (IoT) provides ubiquitous connectivity services to remote areas, oceans, and aviation scenarios via LEO satellites. Applying ISAC technology to satellite IoT systems can simultaneously perform sensing tasks such as target positioning and environmental monitoring while transmitting data, demonstrating significant application value.

[0004] However, satellite communications face severe path loss and channel fading, limiting system performance. Reconfigurable Intelligent Surface (RIS) technology can effectively improve satellite link quality and compensate for signal propagation loss.

[0005] Although RIS-assisted and ISAC technologies have significant application value in satellite IoT systems, existing research on RIS-assisted ISAC mainly focuses on terrestrial communication scenarios. Resource allocation methods for RIS-assisted ISAC in the specific scenario of satellite IoT have not been fully studied. Furthermore, there is currently a lack of a unified optimization framework that can simultaneously support both D-RIS (traditional diagonal RIS) and BD-RIS (super-diagonal RIS) architectures. This prevents the full utilization of the technical advantages of different RIS architectures and makes it difficult to adapt to the complex and ever-changing communication and sensing requirements of satellite IoT scenarios, thus limiting the large-scale application and performance optimization of RIS-assisted ISAC technology in satellite IoT systems. Summary of the Invention

[0006] Therefore, it is necessary to provide a RIS-assisted satellite Internet of Things (ISAC) multi-domain resource scheduling method and apparatus to address the aforementioned technical problems.

[0007] A RIS-assisted satellite IoT ISAC multi-domain resource scheduling method includes the following steps: Construct a RIS-assisted satellite Internet of Things (ISAC) system, the system including configuration N t LEO satellite launch node with root antenna, configuration N RIS of a reflection unit K A single-antenna IoT user receiving node and multiple sensing targets; the RIS is a traditional diagonal RIS or a super-diagonal RIS; Construct channel models from the satellite transmitter node to the RIS, from the RIS to the IoT user receiver node, and from the satellite transmitter node to the IoT user receiver node, and obtain the channel state information corresponding to the channel models; Based on the channel state information, the sensing equivalent channel matrix is ​​calculated, and communication performance indicators and sensing performance indicators are constructed to maximize the overall communication and sensing performance and construct an optimization problem. The optimization problem is decomposed into a beamforming optimization subproblem and a RIS phase shift optimization subproblem using an alternating optimization framework. The RIS phase shift optimization subproblem is either a traditional diagonal RIS phase shift optimization subproblem or a super-diagonal RIS phase shift optimization subproblem. The beamforming optimization subproblem is solved using a weighted minimum mean square error algorithm, the traditional diagonal RIS phase shift optimization subproblem is solved using a gradient-based optimization method, and the super-diagonal RIS phase shift optimization subproblem is solved using a Riemann gradient descent combined with a Givens rotation search strategy. The beamforming matrix and RIS reflection coefficient matrix are output to complete the integrated communication and sensing transmission.

[0008] In one embodiment, the channel model employs the Rician fading model.

[0009] In one embodiment, the communication performance metric uses the user achievable rate and is calculated according to the following formula:

[0010] in, For the first k The achievable rate for each user; For the first k The dryness ratio of information per user; The perception performance index uses perception mutual information and is calculated according to the following formula:

[0011] in, For mutual information perception; det(·) is the matrix determinant; for L 3D identity matrix; To sense noise power; To perceive the equivalent channel matrix; Let be the covariance matrix of the transmitted signal.

[0012] In one embodiment, the optimization problem is:

[0013] in, Represents beamforming matrix and RIS reflection coefficient matrix To maximize the optimization variables; The objective function value; This serves as a weighting factor between communication and sensing. Normalized parameters for perceptual mutual information; tr(·) is the communication rate normalization parameter; tr(·) is the trace of the matrix; This is the satellite's maximum transmission power; This is the feasible region for RIS.

[0014] In one embodiment, for a conventional diagonal RIS:

[0015] in, For the first n The phase of each RIS reflector unit; j The imaginary unit; For superdiagonal RIS:

[0016] Where H is the Hermitian transpose; for N 3D identity matrix.

[0017] In one embodiment, the weighted least mean square error algorithm is used to solve the beamforming optimization subproblem, including: Calculate the first k MMSE receiver filter for each user:

[0018] in, For the first k MMSE receiver filter for each user; For the first k Hermitian transpose of the equivalent channel vector for each user; For the first k Beamforming vectors for each user; For complex conjugate; For the first m Beamforming vectors for each user ;|·| 2 The square of the modulus of the complex number; Communication noise power; Calculate the first k Mean squared error for each user:

[0019] in, For the first k The mean squared error for each user; Re(·) is the real part extraction operation; Update the first based on mean square error k Weighted minimum mean square error matrix for each user:

[0020] in, For the first k The weighted minimum mean square error weight matrix for each user.

[0021] In one embodiment, a gradient-based optimization method is used to solve the traditional diagonal RIS phase shift optimization subproblem, including: Calculate the objective function value Regarding the first n Phase of each RIS reflector unit gradient:

[0022] in, For the objective function with respect to the th n The gradient of each phase; Weighting coefficients for the communication target; Weighting coefficients for the perceived target; Based on gradient Update phase :

[0023] Here, arg(·) is the operation for taking the argument of a complex number.

[0024] In one embodiment, a Riemann gradient descent combined with a Givens rotation search strategy is used to solve the superdiagonal RIS phase shift optimization subproblem, including: Calculate the objective function value The gradient matrix in Euclidean space is obtained by projecting it onto the tangent space, thus obtaining the gradient matrix of the tangent space:

[0025] in, The gradient matrix of the tangent space; Let be the gradient matrix in Euclidean space; Update the RIS reflection coefficient matrix along the tangent space direction:

[0026] in, This is the updated RIS reflection coefficient matrix; This is a singular value decomposition operation; It is a left singular vector matrix; It is a right singular vector matrix; It is a singular value diagonal matrix; This is the gradient descent step size; The RIS reflection coefficient matrix is ​​updated by performing a local fine-tuning search using Givens rotation:

[0027] in, Here is a Givens rotation matrix, which acts on the first... i , l OK, , Givens rotation angle, For Givens rotation phase parameters.

[0028] In one embodiment, the perceived equivalent channel matrix is ​​calculated according to the following formula:

[0029] in, Here is the channel matrix from RIS to the sensing target; The channel matrix from the satellite launch node to the RIS; This is the RIS reflection coefficient matrix.

[0030] A RIS-assisted satellite Internet of Things (ISAC) multi-domain resource scheduling device includes: A model building module is used to build a RIS-assisted satellite Internet of Things (ISAC) system, the system including configuration. N t LEO satellite launch node with root antenna, configuration N RIS of a reflection unit K A single-antenna IoT user receiving node and multiple sensing targets; the RIS is a traditional diagonal RIS or a super-diagonal RIS; The channel construction module is used to construct channel models from the satellite transmitting node to the RIS, from the RIS to the IoT user receiving node, and from the satellite transmitting node to the IoT user receiving node, and to obtain the channel state information corresponding to the channel models. The optimization problem construction module is used to calculate the sensing equivalent channel matrix based on the channel state information, construct communication performance indicators and sensing performance indicators, and construct an optimization problem to maximize the overall communication and sensing performance. The problem decomposition module is used to decompose the optimization problem into beamforming optimization sub-problems and RIS phase shift optimization sub-problems using an alternating optimization framework. The RIS phase shift optimization sub-problems are either traditional diagonal RIS phase shift optimization sub-problems or super-diagonal RIS phase shift optimization sub-problems. The output module is used to solve the beamforming optimization subproblem using the weighted minimum mean square error algorithm, the traditional diagonal RIS phase shift optimization subproblem using the gradient-based optimization method, and the super-diagonal RIS phase shift optimization subproblem using Riemann gradient descent combined with Givens rotation search strategy; it outputs the beamforming matrix and the RIS reflection coefficient matrix to complete the integrated communication and sensing transmission.

[0031] The aforementioned RIS-assisted satellite IoT ISAC multi-domain resource scheduling and device decomposes the optimization problem into beamforming optimization sub-problems and RIS phase shift optimization sub-problems through an alternating optimization framework. Different solution strategies are employed for different RIS types, while a unified equivalent channel model is used. This allows the constructed RIS-assisted satellite IoT ISAC system to be compatible with both traditional diagonal RIS and super-diagonal RIS, providing a unified system model to fully leverage different RIS architectures and adapt to the complex and ever-changing communication and sensing requirements of satellite IoT scenarios. For the super-diagonal RIS phase shift optimization sub-problem, a Riemann gradient descent combined with a Givens rotation search strategy is used to solve the problem, transforming the non-convex constraint problem in Euclidean space into an unconstrained optimization problem on a Riemannian manifold, significantly improving solution efficiency and optimization accuracy. By jointly optimizing beamforming and RIS phase shift, the integrated communication and sensing performance is maximized, resulting in a significant performance improvement compared to schemes that only optimize beamforming or random phase. Attached Figure Description

[0032] Figure 1 This is a flowchart illustrating the RIS-assisted satellite IoT ISAC multi-domain resource scheduling method in one embodiment; Figure 2 This is a schematic diagram of a RIS-assisted satellite Internet of Things (ISAC) system model in one embodiment; Figure 3 This is a schematic diagram comparing the D-RIS and BD-RIS architectures in one embodiment; Figure 4 This is a schematic diagram of the alternating optimization algorithm in one embodiment; Figure 5 This is a structural block diagram of the RIS-assisted satellite IoT ISAC multi-domain resource scheduling device in one embodiment; Figure 6This is a simulation comparison graph showing the system performance under different RIS architectures as a function of the number of RIS reflection units in one embodiment. Figure 7 This is a simulation comparison graph showing the performance of the system under different RIS architectures as a function of transmit power in one embodiment. Detailed Implementation

[0033] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0034] In one embodiment, such as Figure 1 , Figure 2 As shown, a RIS-assisted satellite IoT ISAC multi-domain resource scheduling method is provided, including the following steps: Step 201: Construct a RIS-assisted satellite Internet of Things (ISAC) system, the system including configuration N t LEO satellite launch node with root antenna, configuration N RIS of a reflection unit K A single-antenna IoT user receiving node and multiple sensing targets; the RIS is a traditional diagonal RIS or a super-diagonal RIS.

[0035] It should be noted that the configuration N t The LEO satellite launch node with each antenna serves simultaneously K A single-antenna IoT user receiving node, and senses L Dimensional goals. Configuration. N The RIS of each reflector unit is deployed between the satellite and the ground to enhance signal transmission, and its deployment can be achieved by being carried by a drone. K A single-antenna IoT user receiving node receives satellite downlink communication signals. The sensing target is what the system needs to sense. L Dimensional goals. Among them, N t , N , K , L All are integers.

[0036] Step 202: Construct channel models from the satellite transmitting node to the RIS, from the RIS to the IoT user receiving node, and from the satellite transmitting node to the IoT user receiving node, and obtain the channel state information corresponding to the channel models.

[0037] Step 203: Calculate the sensing equivalent channel matrix based on the channel state information, and construct communication performance indicators and sensing performance indicators to maximize the overall communication and sensing performance to solve the optimization problem.

[0038] It should be noted that by calculating the perceptual equivalent channel matrix, a unified equivalent channel expression is achieved, which makes the expressions of traditional diagonal RIS and superdiagonal RIS under different constraints mathematically unified.

[0039] Step 204: The optimization problem is decomposed into beamforming optimization subproblems and RIS phase shift optimization subproblems using an alternating optimization framework. The RIS phase shift optimization subproblems are either traditional diagonal RIS phase shift optimization subproblems or super-diagonal RIS phase shift optimization subproblems.

[0040] Step 205: The beamforming optimization subproblem is solved using the weighted minimum mean square error algorithm, the traditional diagonal RIS phase shift optimization subproblem is solved using the gradient-based optimization method, and the super-diagonal RIS phase shift optimization subproblem is solved using Riemann gradient descent combined with Givens rotation search strategy; the beamforming matrix and RIS reflection coefficient matrix are output to complete the integrated communication and sensing transmission.

[0041] In the aforementioned RIS-assisted satellite IoT ISAC multi-domain resource scheduling method, the optimization problem is decomposed into beamforming optimization sub-problems and RIS phase shift optimization sub-problems through an alternating optimization framework. Different solution strategies are adopted for different RIS types, while a unified equivalent channel model is performed. This allows the constructed RIS-assisted satellite IoT ISAC system to be compatible with traditional diagonal RIS and super-diagonal RIS, providing a unified system model to fully leverage different RIS architectures and adapt to the complex and ever-changing communication and sensing requirements of satellite IoT scenarios. For the super-diagonal RIS phase shift optimization sub-problem, a Riemann gradient descent combined with a Givens rotation search strategy is used to solve it, transforming the non-convex constraint problem in Euclidean space into an unconstrained optimization problem on a Riemannian manifold, significantly improving solution efficiency and optimization accuracy. By jointly optimizing beamforming and RIS phase shift, the integrated communication and sensing performance is maximized, resulting in a significant performance improvement compared to schemes that only optimize beamforming or random phase.

[0042] In one embodiment, the channel model employs the Rician fading model.

[0043] In this embodiment, considering the characteristics of LEO satellites such as large path loss and Doppler frequency shift, a Rician fading model is adopted to make the channel model independent of the specific architecture of the RIS, and applicable to both diagonal and off-diagonal RIS.

[0044] It should be noted that the Rician fading model includes both line-of-sight (LoS) and non-line-of-sight (NLoS) components.

[0045] In step 202, the channel matrix from the satellite transmitting node (hereinafter referred to as the satellite) to the RIS... for:

[0046] in, The path loss of the link from the satellite launch node to the RIS; The Rician factor for the satellite launch node to the RIS link reflects the relative intensity of the line-of-sight component and the non-line-of-sight component. For deterministic line-of-sight channel components, ; For random non-line-of-sight channel components, .

[0047] RIS to the k Channel vector of each user (IoT user receiving node, hereinafter referred to as user) for:

[0048] in, For RIS to the 1st k Path loss per user ; Rician factor for the RIS-to-user link; For line-of-sight channel components; This refers to the non-line-of-sight channel component.

[0049] Satellite to the k Direct link channel vector of each user for:

[0050] in, For the satellite to the k Direct link path loss for each user; Rician factor for direct links; For line-of-sight channel components; This refers to the non-line-of-sight channel component.

[0051] Satellite transmission signal vector for:

[0052] in, Beamforming matrix , For the firstk Beams for individual users; To send the symbol vector, , For the first k The sent symbols of each user satisfy the following conditions. , Let be the mathematical expectation.

[0053] No. k Equivalent channel vector for each user for:

[0054] in, For RIS to the 1st k Channel matrix for each user ; This is the RIS reflection coefficient matrix, which is understandable for D-RIS. For BD-RIS, .

[0055] The received signal of the kth user for:

[0056] in, For the first k Hermitian transpose of the equivalent channel vector for each user; For the first m Beamforming vectors for each user; For the first m Sending symbols for each user; For the first k Additive white Gaussian noise at each user location , This represents the communication noise power.

[0057] No. k The signal-to-interference-plus-noise ratio for each user is: .

[0058] The constructed RIS-assisted satellite IoT ISAC system can support both traditional diagonal RIS (D-RIS) and super-diagonal RIS (BD-RIS). (See reference...) Figure 3 For D-RIS, the reflection coefficient matrix of D-RIS It is a diagonal matrix, and each reflecting unit independently controls the phase:

[0059] Where diag(·) is a diagonal matrix; For the firstn The phase of each RIS reflector unit, , ; j The imaginary unit; For the first n The reflection coefficient of each reflecting unit satisfies the unit modulus constraint. .

[0060] For BD-RIS, the reflection coefficient matrix of BD-RIS It is a full matrix that satisfies symmetric unitary constraints, allowing coupling between reflection elements:

[0061] in, for N 3D identity matrix; superscript T denotes matrix transpose, superscript H denotes Hermitian transpose (conjugate transpose); symmetry constraints This ensures that the incident and reflected signals follow the same phase response; unitary constraint. This ensures lossless reflection, meaning the reflected energy equals the incident energy.

[0062] In one embodiment, the communication performance metric uses the user achievable rate and is calculated according to the following formula:

[0063] in, For the first k The achievable rate for each user; For the first k The dryness ratio of information per user; The perception performance index uses perception mutual information and is calculated according to the following formula:

[0064] in, For mutual information perception; det(·) is the matrix determinant; for L 3D identity matrix; To sense noise power; To perceive the equivalent channel matrix, ; Let be the covariance matrix of the transmitted signal.

[0065] In this embodiment, perceptual mutual information is used as a radar performance indicator, which can more comprehensively characterize perception performance.

[0066] In one embodiment, the perceived equivalent channel matrix is ​​calculated according to the following formula:

[0067] in, Here is the channel matrix from RIS to the sensing target; The channel matrix from the satellite launch node to the RIS; This is the RIS reflection coefficient matrix.

[0068] In one embodiment, the optimization problem is:

[0069] in, Represents beamforming matrix and RIS reflection coefficient matrix To maximize the optimization variables; The objective function value; This serves as a weighting factor between communication and sensing. Normalized parameters for perceptual mutual information; tr(·) is the communication rate normalization parameter; tr(·) is the trace of the matrix; This is the satellite's maximum transmission power; This represents the feasible region of RIS. Constraint C1 represents the satellite transmit power constraint, constraint C2 represents the user quality of service (QoS) constraint, and constraint C3 represents the RIS reflection coefficient constraint.

[0070] Considering that the optimization problem is a non-convex optimization problem concerning the beamforming matrix and the RIS reflection coefficient matrix, an alternating optimization framework is adopted to decompose the optimization problem into beamforming optimization subproblems and RIS phase shift optimization subproblems, and then iteratively solves them until convergence. The algorithm flow of the alternating optimization framework is as follows: Figure 4 As shown, the specific steps include: S1. Set the initial beamforming matrix Set the initial RIS phase shift matrix Set the convergence threshold ε and the iteration counter. t =0; S2. Fix the current RIS phase shift matrix and solve the subproblem. The beamforming vector is solved iteratively using the weighted least mean square error algorithm. S3, Fixed updated beamforming matrix Solve the subproblems ; S4. Calculate the increment of the objective function. ,like If the algorithm converges, it will output the optimal beamforming matrix. With RIS reflection coefficient matrix and convergence value ,like Update the iteration counter t =t +1, return to S2.

[0071] The alternating optimization framework decomposes the original non-convex optimization problem into two subproblems that are solved iteratively, thereby decoupling the high-dimensional coupled problem and reducing the difficulty of solving it; at the same time, each subproblem has an independently designed algorithm. The alternating optimization framework guarantees convergence, and the objective function remains monotonically constant.

[0072] In one embodiment, for a conventional diagonal RIS:

[0073] in, For the first n The phase of each RIS reflector unit; j The imaginary unit; For superdiagonal RIS:

[0074] Where H is the Hermitian transpose; for N 3D identity matrix.

[0075] In this embodiment, for different RIS types, the RIS feasible domain needs to satisfy the characteristics corresponding to the RIS type.

[0076] In one embodiment, the weighted least mean square error algorithm is used to solve the beamforming optimization subproblem, including: Calculate the first k MMSE receiver filter for each user:

[0077] in, For the first k MMSE receiver filter for each user; For the first k Hermitian transpose of the equivalent channel vector for each user; For the first k Beamforming vectors for each user; For complex conjugate; For the first m Beamforming vectors for each user ;|·| 2 The square of the modulus of the complex number; Communication noise power; Calculate the first k Mean squared error for each user:

[0078] in, For the firstk The mean squared error for each user; Re(·) is the real part extraction operation; Update the first based on mean square error k Weighted minimum mean square error matrix for each user:

[0079] in, For the first k The weighted minimum mean square error weight matrix for each user.

[0080] It is important to note that That is, there is a logarithmic relationship between the rate and the minimum MSE.

[0081] In this embodiment, the beamforming optimization subproblem is solved by using the weighted minimum mean square error algorithm, taking advantage of the equivalence between rate maximization and MSE (mean square error) minimization.

[0082] In one embodiment, a gradient-based optimization method is used to solve the traditional diagonal RIS phase shift optimization subproblem, including: Calculate the objective function value Regarding the first n Phase of each RIS reflector unit gradient:

[0083] in, For the objective function with respect to the th n The gradient of each phase; Weighting coefficients for the communication target; Weighting coefficients for the perceived target; Represents the total communication rate Regarding the first n Phase The partial derivatives; Represents perceived mutual information Regarding the first n Phase The partial derivatives of .

[0084] Based on gradient Update phase :

[0085] Here, arg(·) is the operation for taking the argument of a complex number.

[0086] In one embodiment, a Riemann gradient descent combined with a Givens rotation search strategy is used to solve the superdiagonal RIS phase shift optimization subproblem, including: Calculate the objective function value The gradient matrix in Euclidean space is obtained by projecting it onto the tangent space, thus obtaining the gradient matrix of the tangent space:

[0087] in, The gradient matrix of the tangent space; Let be the gradient matrix in Euclidean space; This represents the unhermitianization operation, used to project the gradient onto the tangent space of a symmetric unitary manifold.

[0088] Update the RIS reflection coefficient matrix along the tangent space direction:

[0089] in, This is the updated RIS reflection coefficient matrix; This is a singular value decomposition operation; It is a left singular vector matrix; It is a right singular vector matrix; It is a singular value diagonal matrix; This is the gradient descent step size; SVD shrinkage ensures that the resulting matrix still satisfies the unitary constraint.

[0090] The RIS reflection coefficient matrix is ​​updated by performing a local fine-tuning search using Givens rotation:

[0091] in, Here is a Givens rotation matrix, which acts on the first... i , l OK, , Givens rotation angle, For Givens rotation phase parameters.

[0092] In this embodiment, Givens rotation performs a local fine-grained search while maintaining unitary constraints, thereby improving optimization accuracy.

[0093] In one embodiment, to verify the effectiveness of the method proposed in this invention, system performance was simulated and analyzed. The simulation parameters were set as follows: satellite altitude 600km, carrier frequency 2GHz, system bandwidth 20MHz, noise power spectral density -174dBm / Hz, Rice factor 10dB, and number of users... K =4, number of perceived targets L =2. The comparison schemes include: no RIS scheme, D-RIS scheme, and BD-RIS scheme.

[0094] The results are as follows Figure 6 , Figure 7 As shown, Figure 6 This demonstrates that, with the same transmit power, the achievable rates of radar sensing mutual information and communication vary with the number of RIS reflector elements. N The curve showing the change. From Figure 6 It can be seen from this: The performance of RIS-free solutions is basically unchanged. N The radar sensing mutual information is approximately 0.46 bits / s / Hz, and the communication rate is approximately 0.44 bits / s / Hz, which serves as a benchmark for comparison. The performance of both the D-RIS and BD-RIS schemes increased. N Increase and significantly improve, when N When the signal strength is 81, the radar sensing mutual information of the D-RIS scheme is approximately 10.28 bits / s / Hz, and the communication rate is approximately 10.01 bits / s / Hz. The BD-RIS scheme is consistently superior to the D-RIS scheme, when N At a resolution of 80, the radar sensing mutual information of the BD-RIS scheme is approximately 11.02 bits / s / Hz, and the communication rate is approximately 15.48 bits / s / Hz, representing improvements of approximately 7.20% and 54.65% respectively compared to D-RIS. This is because the super-diagonal structure of BD-RIS enables more flexible beam control and provides higher beamforming gain.

[0095] Figure 7 The curves showing the variation of radar sensing mutual information and achievable communication rate with transmit power are presented when the number of RIS reflector elements is the same. From Figure 7 It can be seen from this: The performance of all three schemes improves with increasing transmit power, consistent with the fundamental laws of information theory. Across the entire power range, the BD-RIS scheme significantly outperforms the D-RIS scheme, while the D-RIS scheme significantly outperforms the scheme without RIS, validating the effectiveness of RIS assistance and the BD-RIS architecture. At a transmit power of 10dB, the BD-RIS scheme achieves approximately 16.38 bit / s / Hz radar sensing mutual information and 18.66 bit / s / Hz communication rate, compared to approximately 10.40 bit / s / Hz and 10.53 bit / s / Hz for the D-RIS scheme, respectively. This represents a performance improvement of approximately 57.50% and 77.21% for BD-RIS compared to D-RIS. In the low-power region (-10dB to -4dB), the advantages of BD-RIS are even more pronounced, which is significant for power-constrained applications in satellite IoT scenarios.

[0096] The simulation results demonstrate that the RIS-assisted satellite IoT ISAC multi-domain resource scheduling method proposed in this invention can effectively improve the system's integrated sensing performance, especially the BD-RIS architecture, which exhibits significant advantages in satellite IoT scenarios. This invention is the first to propose a unified optimization framework that simultaneously supports D-RIS and BD-RIS, and proposes an efficient solution algorithm based on Riemannian manifold optimization and Givens rotation for the symmetric unitary constraints of BD-RIS, filling the gap in existing technologies for BD-RIS-assisted satellite IoT ISAC systems.

[0097] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0098] In one embodiment, such as Figure 5 As shown, a RIS-assisted satellite Internet of Things (ISAC) multi-domain resource scheduling device is provided, comprising: Model building module 901 is used to build a RIS-assisted satellite Internet of Things (ISAC) system, the system including configuration N t LEO satellite launch node with root antenna, configuration N RIS of a reflection unit K A single-antenna IoT user receiving node and multiple sensing targets; the RIS is a traditional diagonal RIS or a super-diagonal RIS.

[0099] The channel construction module 902 is used to construct channel models from the satellite transmitting node to the RIS, from the RIS to the IoT user receiving node, and from the satellite transmitting node to the IoT user receiving node, and to obtain the channel state information corresponding to the channel models.

[0100] The optimization problem construction module 903 is used to calculate the sensing equivalent channel matrix based on the channel state information, construct communication performance indicators and sensing performance indicators, and construct an optimization problem to maximize the overall communication and sensing performance.

[0101] Problem decomposition module 904 is used to decompose the optimization problem into beamforming optimization subproblems and RIS phase shift optimization subproblems using an alternating optimization framework. The RIS phase shift optimization subproblems are either traditional diagonal RIS phase shift optimization subproblems or superdiagonal RIS phase shift optimization subproblems.

[0102] The output module 905 is used to solve the beamforming optimization subproblem using the weighted minimum mean square error algorithm, the traditional diagonal RIS phase shift optimization subproblem using the gradient-based optimization method, and the super-diagonal RIS phase shift optimization subproblem using Riemann gradient descent combined with Givens rotation search strategy; it outputs the beamforming matrix and the RIS reflection coefficient matrix to complete the integrated communication and sensing transmission.

[0103] Specific limitations regarding the RIS-assisted satellite IoT ISAC multi-domain resource scheduling device can be found in the above-mentioned limitations on the RIS-assisted satellite IoT ISAC multi-domain resource scheduling method, and will not be repeated here. Each module in the aforementioned RIS-assisted satellite IoT ISAC multi-domain resource scheduling device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware or independently of the processor in a computer device, or stored in software in the memory of a computer device, so that the processor can call and execute the operations corresponding to each module.

[0104] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0105] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A RIS-assisted satellite Internet of Things (ISAC) multi-domain resource scheduling method, characterized in that, Includes the following steps: Construct a RIS-assisted satellite Internet of Things (ISAC) system, the system including configuration N t LEO satellite launch node with root antenna, configuration N RIS of a reflection unit K A single-antenna IoT user receiving node and multiple sensing targets; The RIS is a traditional diagonal RIS or a superdiagonal RIS; Construct channel models from the satellite transmitter node to the RIS, from the RIS to the IoT user receiver node, and from the satellite transmitter node to the IoT user receiver node, and obtain the channel state information corresponding to the channel models; Based on the channel state information, the sensing equivalent channel matrix is ​​calculated, and communication performance indicators and sensing performance indicators are constructed to maximize the overall communication and sensing performance and construct an optimization problem. The optimization problem is decomposed into a beamforming optimization subproblem and a RIS phase shift optimization subproblem using an alternating optimization framework. The RIS phase shift optimization subproblem is either a traditional diagonal RIS phase shift optimization subproblem or a super-diagonal RIS phase shift optimization subproblem. The beamforming optimization subproblem is solved using a weighted minimum mean square error algorithm, the traditional diagonal RIS phase shift optimization subproblem is solved using a gradient-based optimization method, and the super-diagonal RIS phase shift optimization subproblem is solved using a Riemann gradient descent combined with a Givens rotation search strategy. The beamforming matrix and RIS reflection coefficient matrix are output to complete the integrated communication and sensing transmission.

2. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 1, characterized in that, The channel model adopts the Rician fading model.

3. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 1, characterized in that, The communication performance metrics are based on the user achievable rate and are calculated using the following formula: in, For the first k The achievable rate for each user; For the first k The dryness ratio of information per user; The perception performance index uses perception mutual information and is calculated according to the following formula: in, For mutual information perception; det(·) is the matrix determinant; for L 3D identity matrix; To sense noise power; To perceive the equivalent channel matrix; Let be the covariance matrix of the transmitted signal.

4. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 3, characterized in that, The optimization problem is: in, Represents beamforming matrix and RIS reflection coefficient matrix To maximize the optimization variables; The objective function value; This serves as a weighting factor between communication and sensing. Normalized parameters for perceptual mutual information; tr(·) is the communication rate normalization parameter; tr(·) is the trace of the matrix; This is the satellite's maximum transmission power; This is the feasible region for RIS.

5. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 4, characterized in that, For traditional diagonal RIS: in, For the first n The phase of each RIS reflector unit; j The imaginary unit; For superdiagonal RIS: Where H is the Hermitian transpose; for N 3D identity matrix.

6. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 4, characterized in that, The weighted least mean square error algorithm is used to solve the beamforming optimization subproblem, including: Calculate the first k MMSE receiver filter for each user: in, For the first k MMSE receiver filter for each user; For the first k Hermitian transpose of the equivalent channel vector for each user; For the first k Beamforming vectors for each user; For complex conjugate; For the first m Beamforming vectors for each user ;|·| 2 The square of the modulus of the complex number; Communication noise power; Calculate the first k Mean squared error for each user: in, For the first k The mean squared error for each user; Re(·) is the real part extraction operation; Update the first based on mean square error k Weighted minimum mean square error matrix for each user: in, For the first k The weighted minimum mean square error weight matrix for each user.

7. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 4, characterized in that, A gradient-based optimization method is used to solve the traditional diagonal RIS phase shift optimization subproblem, including: Calculate the objective function value Regarding the first n Phase of each RIS reflector unit gradient: in, For the objective function with respect to the th n The gradient of each phase; Weighting coefficients for the communication target; Weighting coefficients for the perceived target; Based on gradient Update phase : Here, arg(·) is the operation for taking the argument of a complex number.

8. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 4, characterized in that, The superdiagonal RIS phase shift optimization subproblem is solved using a combination of Riemann gradient descent and Givens rotation search strategy, including: Calculate the objective function value The gradient matrix in Euclidean space is obtained by projecting it onto the tangent space, thus obtaining the gradient matrix of the tangent space: in, The gradient matrix of the tangent space; Let be the gradient matrix in Euclidean space; Update the RIS reflection coefficient matrix along the tangent space direction: in, This is the updated RIS reflection coefficient matrix; This is a singular value decomposition operation; It is a left singular vector matrix; It is a right singular vector matrix; It is a singular value diagonal matrix; This is the gradient descent step size; The RIS reflection coefficient matrix is ​​updated by performing a local fine-tuning search using Givens rotation: in, Here is a Givens rotation matrix, which acts on the first... i , l OK, , Givens rotation angle, For Givens rotation phase parameters.

9. The RIS-assisted satellite IoT ISAC multi-domain resource scheduling method according to claim 3, characterized in that, The sensing equivalent channel matrix is ​​calculated according to the following formula: in, Here is the channel matrix from RIS to the sensing target; The channel matrix from the satellite launch node to the RIS; This is the RIS reflection coefficient matrix.

10. A RIS-assisted satellite Internet of Things (ISAC) multi-domain resource scheduling device, characterized in that, include: A model building module is used to build a RIS-assisted satellite Internet of Things (ISAC) system, the system including configuration. N t LEO satellite launch node with root antenna, configuration N RIS of a reflection unit K A single-antenna IoT user receiving node and multiple sensing targets; The RIS is a traditional diagonal RIS or a superdiagonal RIS; The channel construction module is used to construct channel models from the satellite transmitting node to the RIS, from the RIS to the IoT user receiving node, and from the satellite transmitting node to the IoT user receiving node, and to obtain the channel state information corresponding to the channel models. The optimization problem construction module is used to calculate the sensing equivalent channel matrix based on the channel state information, construct communication performance indicators and sensing performance indicators, and construct an optimization problem to maximize the overall communication and sensing performance. The problem decomposition module is used to decompose the optimization problem into beamforming optimization sub-problems and RIS phase shift optimization sub-problems using an alternating optimization framework. The RIS phase shift optimization sub-problems are either traditional diagonal RIS phase shift optimization sub-problems or super-diagonal RIS phase shift optimization sub-problems. The output module is used to solve the beamforming optimization subproblem using the weighted minimum mean square error algorithm, the traditional diagonal RIS phase shift optimization subproblem using the gradient-based optimization method, and the super-diagonal RIS phase shift optimization subproblem using Riemann gradient descent combined with Givens rotation search strategy; it outputs the beamforming matrix and the RIS reflection coefficient matrix to complete the integrated communication and sensing transmission.

Images