A method and device for scheduling narrowband transmission of multimodal data for ocean satellite Internet of Things

Abstract

The present invention provides a method and device for scheduling narrowband transmission of multimodal data of ocean satellite Internet of Things, which belongs to the technical field of ocean satellite Internet of Things. The method innovatively proposes a multimodal data narrowband transmission scheduling strategy to address the challenges of offshore observation equipment in energy-constrained and dynamic satellite channel environments. Its core includes: constructing a dynamic transmission model for multimodal data, and conducting data value evaluation based on spatiotemporal intelligence; designing congestion perception and hierarchical congestion control strategies to ensure data timeliness; optimizing transmission fairness by combining data value weight factors, PAoI and Jain fairness index; constructing an energy dynamic evolution model to ensure long-term energy balance of equipment; based on deep Q network and imitation learning, combined with a secure reinforcement learning framework, jointly optimizing information timeliness, energy consumption and congestion. The present invention significantly reduces the data transmission delay, energy consumption and congestion incidence rate of deep-sea observation and detection, and improves information timeliness and fairness of multimodal data transmission scheduling.

Description

Technical Field

[0001] The present invention relates to the technical field of ocean satellite Internet of Things, and in particular to a method and device for scheduling narrowband transmission of multimodal data of an ocean satellite Internet of Things. Background Art

[0002] The integration of ocean observation and detection equipment with low-orbit satellite internet has become a key infrastructure for deep-sea environmental observation and target detection. To meet the multi-dimensional perception needs of complex ocean environments and targets, drifting buoys, wave energy gliders, Argo buoys, and real-time submersibles are equipped with meteorological sensors, hydrological sensors, and optoelectronic cameras to collect multimodal data such as ocean meteorology, hydrology, and targets. Data backhaul technology based on low-orbit satellite internet has broken through the coverage limitations of traditional shore-based communications and provided wide-area interconnection capabilities for deep-sea areas. However, the contradiction between the narrowband communication characteristics of satellites and the differentiated transmission requirements of multimodal data has become increasingly prominent, making it difficult to guarantee the transmission scheduling efficiency and service quality of multimodal data. How to achieve efficient narrowband transmission scheduling of multimodal data under multimodal data, dynamic network status, and strict energy constraints has become a core issue that needs to be addressed.

[0003] In deep-sea observation and detection scenarios, the transmission requirements of multimodal data show significant temporal and spatial heterogeneity. Periodic meteorological and hydrological data need to be continuously and stably transmitted back to support long-term monitoring of the marine environment; while event-triggered data such as target discovery, hydrological anomalies, and meteorological warnings require low-latency, highly reliable transmission. In addition, due to the influence of satellite communication windows, fluctuations in solar power supply for equipment, and the complex marine environment, data transmission needs to maximize transmission value within a limited communication window. Existing transmission scheduling methods have deficiencies in data value quantification, multi-objective collaborative optimization, and adaptive congestion control, and cannot meet the differentiated transmission needs of multimodal data in complex scenarios.

[0004] Currently, the narrowband transmission of multimodal data in the ocean satellite IoT faces multiple challenges: First, insufficient spatiotemporal feature mining leads to low transmission efficiency. Existing mechanisms lack in-depth analysis of the spatiotemporal correlation characteristics of multimodal data, and data transmission priority allocation lacks a scientific basis. Second, congestion control is difficult under the constraints of satellite link resources. Traditional congestion sensing typically relies on a single metric and lacks multi-dimensional perception of bandwidth utilization, delay jitter, and queue length change rate, making it difficult to adapt to dynamic scenarios such as severe weather and data peaks. Third, the energy supply of deep-sea observation and detection equipment is unstable. Traditional energy acquisition models do not consider the day-night differences in solar power supply, and data transmission does not consider the long-term energy balance of equipment. Therefore, to meet the needs of multimodal data transmission in complex deep-sea environments, it is urgent to design a narrowband transmission scheduling method that integrates spatiotemporal intelligent analysis, adaptive congestion control, and dynamic energy optimization.

[0005] Existing transmission scheduling strategies are mostly based on a single optimization objective, ignoring the resource competition and coupling relationship between information timeliness, energy consumption and congestion. Traditional methods may reduce energy consumption by sacrificing information timeliness, or blindly discard data during congestion, resulting in the loss of key information, and are unable to achieve dynamic trade-offs among multiple objectives. The multi-objective joint optimization model proposed in this invention incorporates the Peak Age of Information (PAoI) of the integrated data value weight, device energy consumption and network congestion into a unified objective function, transforming the multimodal data transmission scheduling problem into a Pareto optimal problem of minimizing the weighted sum. The model can adaptively adjust the weight coefficients of each optimization item, balance information freshness, energy efficiency and network load while ensuring the priority of key data transmission, and significantly improve the overall performance of the system.

[0006] When it comes to solving high-dimensional dynamic optimization problems, existing approaches such as greedy strategies based on Lyapunov optimization, heuristic algorithms, and linear programming struggle to handle the nonlinear, time-varying constraints and large state spaces required for multimodal data scheduling in ocean satellite IoT. While traditional reinforcement learning (RL) approaches are capable of handling complex decisions, they are susceptible to training instability in dynamic ocean environments due to low exploration efficiency and high constraint violation risks. This paper innovatively integrates Deep Q-Network (DQN) with imitation learning to construct a secure RL framework. Expert policy pre-training accelerates DQN convergence, while the action probability distribution provided by imitation learning guides exploration. Furthermore, a safety constraint layer is introduced to transform bandwidth and energy constraints into feasible region boundaries in the state space, ensuring that the actions generated by the policy naturally satisfy physical constraints. This hybrid solution not only reduces training costs but also stably generates optimal scheduling strategies that satisfy multi-objective trade-offs in complex and dynamic ocean environments, providing theoretical feasibility and technological advancement for multimodal data transmission in narrowband satellite IoT. Summary of the Invention

[0007] In view of this, an embodiment of the present invention provides a method and device for scheduling narrowband transmission of multimodal data of an ocean satellite Internet of Things to eliminate or improve one or more defects existing in the prior art, and solve the problems of insufficient analysis of the spatiotemporal characteristics of existing multimodal data, limited satellite communication link resources, unstable equipment energy supply, and the fact that existing transmission scheduling methods are not fully applicable to the complex dynamic environment of the ocean satellite Internet of Things.

[0008] In one aspect, the present invention provides a method for scheduling narrowband transmission of multimodal data of an ocean satellite Internet of Things, characterized in that the method comprises the following steps:

[0009] Construct a dynamic transmission model for multimodal data of deep-sea observation and detection equipment: This model receives multimodal data collected by meteorological sensors, hydrological sensors, and optoelectronic cameras, and establishes independent data queues according to the nature of the data (periodic meteorological and hydrological data, event-triggered target discovery, hydrological anomalies, and meteorological warning data), and dynamically updates the queue length.

[0010] Aiming at the bandwidth limitation of satellite narrowband communication, a bandwidth resource constraint model is constructed to ensure that data transmission is carried out efficiently within limited bandwidth resources.

[0011] Carry out data value assessment driven by spatiotemporal intelligence, use the Long Short-Term Memory (LSTM) network to analyze the fluctuation characteristics of the time dimension, combine the density-based spatial clustering of applications with noise (DBSCAN) algorithm to spatially cluster deep-sea hotspots, quantify data value factors, and realize differentiated data scheduling.

[0012] Design a congestion-aware and hierarchical congestion control strategy. Calculate the network congestion index based on bandwidth utilization, latency jitter, and queue length change rate. Dynamically adjust the index weights based on different scenarios (such as severe weather, peak data transmission periods, and sudden data influxes) to achieve scenario-adaptive network congestion assessment. Design a threshold-based hierarchical congestion control strategy to adjust data transmission strategies based on the degree of congestion.

[0013] An information timeliness model is constructed, using the PAoI (Patterning of Index) to evaluate the timeliness of data packets. Combined with the data value weight factor, a weighted PAoI metric is constructed to evaluate the timeliness of data of varying value. A multi-sensor PAoI fairness function, incorporating data value weights, is constructed based on the Jain Fairness Index to ensure fairness in multi-sensor data transmission and prevent high-value data from occupying resources for extended periods.

[0014] Establish an energy dynamic evolution model for deep-sea observation and detection equipment, establish a dynamic energy acquisition model based on the characteristics of solar power supply and changes in the day and night cycle, and design an energy consumption model based on the transmission and sleep energy consumption characteristics of the sensor to ensure that the equipment meets long-term energy balance.

[0015] A multi-objective optimization function is constructed to minimize the weighted sum of weighted PAoI, energy consumption, and congestion. A reinforcement learning strategy combining DQN and imitation learning is employed. The state and action spaces are modeled using a Markov decision process. A safety constraint layer is introduced to limit bandwidth and energy violations, generating a dynamic optimal transmission scheduling strategy.

[0016] In some embodiments of the present invention, the method further comprises:

[0017] The hierarchical congestion control strategy dynamically adjusts weight allocation for different scenarios. For example, it increases the weight of delay jitter in bad weather and increases the weight of queue change rate when burst data is influx, enhancing scenario adaptability.

[0018] The fair transmission scheduling balances the priority transmission of high-value data and the basic service requirements of low-value data by integrating the data value weight with the Jain fairness index, thereby avoiding data starvation.

[0019] The energy acquisition model designs a diurnal cycle adaptive adjustment mechanism to maximize energy storage during the day and optimize transmission tasks at night to reduce energy consumption and extend the equipment life cycle.

[0020] The optimization problem-solving method based on DQN and imitation learning utilizes expert historical data to pre-train the DQN network, combines imitation learning to accelerate convergence, and dynamically adjusts the exploration rate to adapt to changes in the ocean environment. A constraint layer is introduced through safe reinforcement learning to limit the action space to avoid violating bandwidth and energy constraints.

[0021] On the other hand, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things as described in any one of the above items.

[0022] The beneficial effects of the present invention are at least:

[0023] The present invention provides a method and device for scheduling narrowband transmission of multimodal data for an ocean satellite Internet of Things. The method comprises: proposing a multi-objective joint optimization method based on weighted PAoI, energy consumption, and congestion, incorporating PAoI, which integrates data value weights, device energy consumption, and network congestion into a unified objective function, and adaptively adjusting the weight coefficients of each optimization item. While ensuring the priority of key data transmission, the method balances information freshness, energy efficiency, and network load, significantly improving overall system performance; and proposing an optimization problem-solving method based on DQN and imitation learning, integrating it with a secure reinforcement learning framework. This method not only reduces training costs but also stably generates an optimal scheduling strategy that satisfies multi-objective trade-offs in complex and dynamic ocean environments, thus providing theoretical feasibility and technological advancement for multimodal data transmission in narrowband satellite Internet of Things.

[0024] Additional advantages, objects, and features of the present invention will be set forth in part in the following description and will become apparent to those skilled in the art upon examination of the following or may be learned from practice of the present invention. The objects and other advantages of the present invention may be realized and obtained by the structures particularly pointed out in the description and drawings.

[0025] Those skilled in the art will understand that the purposes and advantages that can be achieved by the present invention are not limited to the above specific descriptions, and the above and other purposes that can be achieved by the present invention will be more clearly understood based on the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The drawings described herein are used to provide a further understanding of the present invention, constitute a part of this application, and do not constitute a limitation of the present invention. In the drawings:

[0027] Figure 1 The figure is a schematic diagram of the steps of a method for scheduling narrowband transmission of multimodal data of an ocean satellite Internet of Things in one embodiment of the present invention.

[0028] Figure 2 This is a diagram of a narrowband transmission scenario of ocean observation and detection equipment data based on low-orbit satellite Internet in one embodiment of the present invention.

[0029] Figure 3 This is a technical roadmap for scheduling narrowband transmission of multimodal data of ocean satellite Internet of Things in one embodiment of the present invention.

[0030] Figure 4 The figure is a schematic diagram of the dynamic transmission scheduling process of the ocean satellite Internet of Things based on DQN and imitation learning in one embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. Here, the exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but are not intended to limit the present invention.

[0032] It should also be noted that, in order to avoid obscuring the present invention due to unnecessary details, the accompanying drawings only show structures and / or processing steps closely related to the solutions according to the present invention, while other details that are not closely related to the present invention are omitted.

[0033] It should be emphasized that the term "include / comprises" when used herein refers to the existence of features, elements, steps or components, but does not exclude the existence or addition of one or more other features, elements, steps or components.

[0034] It should also be noted that, unless otherwise specified, the term "connection" herein may refer not only to a direct connection but also to an indirect connection involving an intermediate.

[0035] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. In the accompanying drawings, the same reference numerals represent the same or similar components, or the same or similar steps.

[0036] In order to solve the problems of insufficient analysis of spatiotemporal characteristics of existing multimodal data, limited satellite communication link resources, unstable equipment energy supply, and the inability of existing transmission scheduling methods to fully apply to the complex dynamic environment of marine satellite Internet of Things, the present invention provides a narrowband transmission scheduling method for multimodal data of marine satellite Internet of Things, such as Figure 1 As shown, the method includes the following steps S101 to S106:

[0037] Step S101: Construct a dynamic transmission model for multimodal data of deep-sea observation and detection equipment. The model input includes multimodal data collected by meteorological sensors, hydrological sensors and target detection sensors in the ocean satellite Internet of Things. Independent data queues are established for periodic meteorological and hydrological data and event-triggered target discovery, hydrological anomaly and meteorological warning data, and the queue length is dynamically updated; a bandwidth resource constraint model is constructed to address the bandwidth limitations of satellite narrowband communications; data value assessment driven by spatiotemporal intelligence is carried out, LSTM is used to analyze the fluctuation characteristics of the time dimension, and DBSCAN is combined to perform spatial clustering of deep-sea hotspot areas, quantify data value factors, and realize data differentiated scheduling.

[0038] Step S102: Design a congestion perception and hierarchical congestion control strategy, comprehensively calculate the network congestion index based on bandwidth utilization, delay jitter, and queue length change rate, dynamically adjust the index weights for severe weather, data transmission peak periods, and sudden data influx scenarios, and implement scenario-adaptive network congestion assessment; design a threshold-based hierarchical congestion control strategy, transmit according to data value in low congestion mode, limit the single transmission volume and give priority to high-value data in medium congestion mode, and suspend low-value data transmission and discard expired data in high congestion mode.

[0039] Step S103: Information timeliness model and fair transmission scheduling. The Age of Information (AoI) is the time difference between the time a data packet is generated and the time it is successfully received. The PAoI is the maximum AoI value within the data packet transmission period. The data value weight factor is combined with the PAoI to construct a weighted PAoI indicator to evaluate the information timeliness of data of different values. Based on the Jain fairness index, a multi-sensor PAoI fairness function that integrates data value weights is constructed to ensure the fairness of multi-sensor data transmission.

[0040] Step S104: Establish an energy dynamic evolution model for deep-sea observation and detection equipment. By combining the characteristics of solar power supply and changes in the day and night cycle, a dynamic energy acquisition model is established. Based on the transmission and sleep energy consumption characteristics of the sensor, an energy consumption model is designed. Battery energy constraints are introduced, the remaining energy is dynamically updated, and the energy fluctuation range is limited to ensure that the equipment meets the long-term energy balance.

[0041] Step S105: Jointly optimize information timeliness, energy consumption, and congestion, construct a multi-objective optimization function, and minimize the weighted sum of weighted PAoI, energy consumption, and congestion.

[0042] Step S106: Adopt a strategy combining DQN and imitation learning, introduce security reinforcement learning, model the state space, action space, security constraints and reward function through the Markov decision process, generate the optimal transmission decision including the sensor transmission decision variables and the amount of transmission data for each time slot, and realize dynamic transmission scheduling.

[0043] In step S101, based on the data generation rules of the sensors of the ocean observation and detection equipment, an independent data queue is established for each sensor; the data queue length is dynamically updated according to the arrival and transmission process of the data packet; a bandwidth resource constraint model is constructed based on the bandwidth limitation of the satellite narrowband communication module; and a data value assessment based on spatiotemporal intelligence is carried out to provide a basis for differentiated data scheduling.

[0044] In the multi-sensor data queue model, based on the data generation rules of the ocean observation and detection equipment sensors, an independent data queue matching the number of sensors is established; periodic meteorological and hydrological observation data are generated at fixed time intervals, and the arrival time is Obey uniform distribution; event-triggered target discovery, hydrological anomaly and meteorological warning data generate data packets according to the Poisson process, and the arrival volume The obedience parameter is Poisson distribution.

[0045] In the dynamic update model of each sensor queue, the formula for dynamically updating the data queue length is as follows according to the arrival and transmission process of the data packet: ,in, For sensors At the moment The queue length, is the transmission decision variable, The amount of data transmitted in a single time, is the amount of newly arrived data, The maximum queue length to avoid storage overflow.

[0046] In the bandwidth resource constraint model, Figure 2 As shown, drifting buoys, wave energy gliders, Argo buoys and real-time submersible buoys are equipped with satellite narrowband communication modules. The upper limit of the total bandwidth resource in a single time slot is ,sensor The bandwidth requirement for a single transmission is , the narrowband communication bandwidth constraint formula is constructed as .

[0047] In the data value assessment model based on spatiotemporal intelligence, the value weight factor of each sensor data is calculated based on the sensor data type, time characteristics and spatial characteristics. The data value analysis based on spatiotemporal intelligence is adopted. In the time dimension, the fluctuation amplitude of the ocean environment data in the time dimension is perceived based on LSTM, abnormal data changes are identified, and the ocean environment fluctuation rating is obtained. , the corresponding time weighting coefficient is In the spatial dimension, DBSCAN is used to cluster the spatial hotspots of high-precision navigation and positioning data, and regional division and pattern recognition of maritime space are performed to achieve spatial hotspot rating. , the corresponding spatial weighting coefficient is ;Through dynamic data value weighting factors , achieving differentiated service quality assurance. For example, if LSTM detects that the water temperature rises by 2°C within 1 hour, the environmental fluctuation rating in the time dimension , the corresponding time weighting coefficient is ; 10 ships were detected gathering in a certain area, marked as hotspots by DBSCAN, and the hotspot area rating in the spatial dimension , the corresponding time weighting coefficient is ; Calculate data value weight factor .

[0048] In step S102, bandwidth utilization, delay jitter, and queue length change rate are monitored in real time to build a network congestion perception model, calculate the network congestion index, dynamically adjust the indicator weights to adapt to the ocean satellite Internet of Things scenario, and control the transmission strategy according to the threshold level.

[0049] In the network congestion perception model, the degree of congestion is dynamically quantified by real-time monitoring of bandwidth utilization, delay jitter, and queue length change rate.

[0050] The satellite link bandwidth utilization is , reflects the current channel load. For example, when the bandwidth utilization reaches 80%, it indicates that the channel is close to saturation.

[0051] The difference between the current round-trip time (RTT) and the minimum RTT represents the intensity of network fluctuations. , effectively identify the unstable state of the link, the maximum delay jitter threshold allowed by the network is , exemplary, inclement weather It may surge from 10ms to 200ms.

[0052] The queue length change rate is , the weighted average of the change rate of each sensor queue length can be obtained , characterizing the queue change trend.

[0053] Calculate the congestion index based on three indicators: bandwidth utilization, delay jitter, and queue length change rate. , ,Exemplary ,Bandwidth Utilization Weight , delay jitter weight , queue change rate weight .

[0054] In some embodiments, the weight is dynamically adjusted according to the characteristics of the ocean environment: in severe weather, atmospheric attenuation causes signal quality to degrade, delay jitter is more sensitive, and priority response is required for link instability, so the weight is increased. Weight; During the peak period of data transmission, the channel load is close to saturation, and bandwidth allocation needs to be strictly controlled to avoid congestion collapse, so the Weight; when sudden data flows in, the queue grows rapidly, and it is necessary to give priority to relieving the queue pressure and preventing overflow, so increase Weight.

[0055] In the threshold-based hierarchical congestion control strategy, the real-time congestion level is , for example, two congestion thresholds are designed and ;when When low congestion mode is used, the data value factor Determines data transmission priority and allows maximum transmission volume ;when , medium congestion model, start traffic shaping, limit the upper limit of single transmission data volume, and give priority to transmitting high-value data ;when When the traffic is in high congestion mode, low-value data transmission is suspended. , if the queue length , the queue is close to overflow, and the oldest data packet is discarded according to timeliness; in the total time statistics period The long-term average congestion index of the system is .

[0056] In step S103, the data value weight and PAoI are combined to construct a weighted PAoI to quantify the timeliness of multimodal data. Based on the Jain fairness index, multi-sensor resource allocation is constrained, and a multi-sensor data fairness transmission scheduling mechanism is designed to balance the priority transmission of high-value data and the basic service requirements of low-value data.

[0057] In the information timeliness modeling, the timeliness differences of data of different values ​​are quantified, and low-latency transmission of high-value data is prioritized; Age of Information (AoI) is an indicator for measuring the timeliness of information. Sensors No. The peak information age PAoI of a data packet is recorded as ,in, is the transmission completion time, The data generation time; there are Sensors are counted to calculate the average PAoI of the system to reflect the overall timeliness. Data value weight Combined with PAoI, the system average weighted PAoI is defined as ,in, The total number of packets transmitted.

[0058] In the multi-sensor data fairness transmission scheduling mechanism, multiple sensors collect data at different periods, data volumes, and data values. In order to quantify timeliness fairness, the Jain fairness index is improved by combining the data value weight factor, and a multi-sensor information timeliness fairness function is constructed. The formula is: , to avoid high-value data from monopolizing resources for a long time and to ensure the basic transmission needs of low-value data; represents the timeliness fairness threshold, when This indicates that The timeliness of the updated information of each sensor is in an unfair state, otherwise, The timeliness of the sensor update information is in a fair state; for example, the system fairness threshold is set ,like ,If the multi-sensor transmission does not satisfy the fairness constraint, resources will be ,forced to be allocated to low-priority sensors.

[0059] In step S104, since the deep-sea observation and detection equipment can only rely on renewable green energy to operate, and the complex environment at sea cannot always guarantee sufficient and stable energy, the equipment has strict requirements on energy consumption. The deep-sea observation and detection equipment energy dynamic evolution model quantifies the direct impact of data transmission scheduling on the survivability of the equipment by coupling energy acquisition randomness, multi-mode energy consumption and battery energy storage hard constraints, and provides an energy dimension decision boundary for the joint optimization of weighted PAoI, energy consumption and congestion; an energy acquisition model is established based on solar power supply and day and night cycle, combined with the transmission / sleep energy consumption model, and under the battery energy constraint, a dynamic energy management mechanism of the equipment is designed to ensure long-term energy balance.

[0060] In the energy acquisition model, the dynamic energy acquisition process of solar power supply is modeled to adapt to the diurnal cycle and seasonal changes. Modeling the energy harvesting process for an exponential distribution of parameters ; According to the day and night cycle Adjust the energy acquisition model, The length of daylight, which is adjusted dynamically with the seasons ; Exemplarily, the rate of energy acquisition during the day follows Exponential distribution, the length of daylight in summer is 14 hours, and the length of daylight in winter is 10 hours. The specific situation is dynamically calibrated according to the geographical location. The seasonal correction factor is: summer , the total charging energy is increased by 20%, in winter , the total charging energy is reduced by 15%.

[0061] In the energy consumption model, the energy consumption of sensor transmission and sleep is quantified, that is, sensor Single transmission consumption and sleep energy consumption satisfy , in the total time statistical period The long-term average energy consumption of the system is .

[0062] The battery energy constraint defines the upper and lower limits of the battery capacity. and , dynamically update the remaining energy , prevent battery overcharge / over discharge and extend device life.

[0063] In the dynamic energy management mechanism of the equipment, taking into account the battery life and stability of deep-sea observation and detection equipment, in order to achieve long-term stable operation under energy fluctuations, the equipment must meet the long-term energy balance constraint, give priority to high-energy transmission tasks during the day, and limit to low-energy operations at night; for example, if , forced into low power mode and suspending data transmission.

[0064] like Figure 3 As shown, steps S102 to S104 are specific steps for constructing a model of a multimodal data narrowband transmission scheduling system for the ocean satellite Internet of Things, and steps S105 and S106 are specific steps for constructing and solving a multi-objective joint optimization problem.

[0065] In step S105 , a multi-objective optimization problem of jointly optimizing weighted PAoI, energy consumption, and congestion is constructed.

[0066] In the construction of the joint optimization objective function, in the scenario where deep-sea observation and detection equipment performs data backhaul based on low-orbit satellite Internet, under the constraints of narrowband communication bandwidth, dynamic energy acquisition and battery capacity, the optimal transmission scheduling scheme for timely backhaul of high-value data, sustainable operation of equipment energy and adaptive network congestion control is sought, balancing weighted PAoI, energy consumption and congestion, and constructing a multi-objective joint optimization problem. The objective function is to minimize the weighted sum of the long-term time average system weighted PAoI, energy consumption and congestion. The formula is: ,in, is the energy consumption penalty coefficient, is the congestion penalty coefficient, which is used to adjust the balance between weighted PAoI, energy consumption and congestion; the scheduling strategy is used to determine the sensor When the updated information of The design of the system needs to balance the immediate timeliness benefits, long-term system energy efficiency and network congestion control under multiple constraints.

[0067] In step S106, DQN and imitation learning are used to generate the optimal dynamic transmission decision under security constraints.

[0068] In the optimization problem solving method based on DQN and imitation learning, the data transmission scheduling problem is modeled as a Markov decision process. The intelligent agent perceives the current state and performs actions by interacting with the environment. The state space includes the sensor queue length, the current congestion index, the current device energy state, the data value weight factor and the weighted PAoI. The action space includes the transmission decision variables And the amount of data transferred in a single time Combined with secure reinforcement learning, a security constraint layer is introduced to constrain actions, including bandwidth constraints and battery energy constraints. The reward function is the negative value of the joint optimization objective to minimize weighted PAoI, energy consumption, and congestion penalty.

[0069] In the combination strategy of DQN and imitation learning, the training efficiency and strategy security are improved by integrating expert experience and autonomous exploration, such as Figure 4 As shown in the figure, the following steps are included: collecting historical expert transmission scheduling data and constructing an expert dataset containing state-action pairs; using supervised learning to pre-train the DQN network and initialize the Q-value function parameters so that the model can quickly master the basic decision-making mode; in the online training phase, expert data and real-time interaction data are mixed in proportion and input into the experience replay pool, combining the action probability distribution of imitation learning with the Q-value update of DQN to dynamically balance exploration and utilization; using a decaying exploration rate mechanism, for example, the initial exploration rate is 0.5, and it decays by 0.1 every 1000 steps, gradually reducing the random exploration ratio, so that the model transitions from relying on expert experience to autonomous optimization, and finally generates an efficient and safe transmission strategy in a complex dynamic ocean environment.

[0070] Safety reinforcement learning ensures that transmission decisions comply with physical limitations and system stability requirements by introducing a constraint layer and penalty mechanism. In the action selection stage, a two-level filtering mechanism is designed: the first level excludes excessive action combinations based on bandwidth constraints; the second level limits risky operations based on energy constraints. At the same time, a violation penalty term is embedded in the reward function, and additional reward values ​​are deducted when the constraints are violated. Negative feedback guides the agent to learn safe actions. For example, when the remaining energy of the device is below When the system is running, the security layer only allows low-energy operations to avoid device downtime. This mechanism is effective throughout the training and inference phases, ensuring reliable operation of the system in narrowband communication and energy-constrained environments.

[0071] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things.

[0072] Corresponding to the above method, the present invention also provides a device, which includes a computer device, the computer device includes a processor and a memory, the memory stores computer instructions, the processor is used to execute the computer instructions stored in the memory, and when the computer instructions are executed by the processor, the device implements the steps of the method described above.

[0073] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program implements the steps of the aforementioned edge computing server deployment method. The computer-readable storage medium can be a tangible storage medium, such as a random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, floppy disk, hard disk, removable storage disk, CD-ROM, or any other form of storage medium known in the art.

[0074] In summary, the present invention provides a method and device for scheduling narrowband transmission of multimodal data of an ocean satellite Internet of Things, including: proposing a multi-objective joint optimization method based on weighted PAoI, energy consumption and congestion, incorporating PAoI, equipment energy consumption and network congestion that integrate data value weights into a unified objective function, adaptively adjusting the weight coefficients of each optimization item, and balancing information freshness, energy efficiency and network load while ensuring the priority of key data transmission, thereby significantly improving the overall performance of the system; proposing an optimization problem solving method based on DQN and imitation learning, integrating a secure reinforcement learning framework, which not only reduces training costs, but also can stably generate an optimal scheduling strategy that meets multi-objective trade-offs in a complex and dynamic ocean environment, providing theoretical feasibility and technological advancement for multimodal data transmission of a narrowband satellite Internet of Things.

[0075] It should be understood by those skilled in the art that the various exemplary components, systems, and methods described in conjunction with the embodiments disclosed herein can be implemented in hardware, software, or a combination of both. Whether to implement the system in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention. When implemented in hardware, it may be, for example, an electronic circuit, an application-specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present invention are programs or code segments used to perform the required tasks. The programs or code segments can be stored in a machine-readable medium or transmitted over a transmission medium or communication link via a data signal carried in a carrier wave.

[0076] It should be understood that the present invention is not limited to the specific configurations and processes described above and illustrated in the figures. For the sake of brevity, a detailed description of known methods is omitted. In the above embodiments, several specific steps are described and illustrated as examples. However, the method of the present invention is not limited to the specific steps described and illustrated. Those skilled in the art may make various changes, modifications, and additions, or change the order of the steps after understanding the spirit of the present invention.

[0077] In the present invention, features described and / or illustrated for one embodiment may be used in the same or similar manner in one or more other embodiments, and / or combined with or replace features of other embodiments.

[0078] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A method for scheduling narrowband transmission of multimodal data of ocean satellite Internet of Things, characterized in that: The method comprises the following steps: A dynamic multimodal data transmission model for deep-sea observation and detection equipment was constructed. Its input included multimodal data collected by meteorological, hydrological, and target monitoring sensors. Independent queues were established for periodic data and event-triggered data, and each queue was dynamically updated. A bandwidth resource constraint model for satellite narrowband communications was constructed. A data value assessment mechanism was implemented that combined time series data feature analysis using a long short-term memory (LSTM) network with spatial clustering based on the density-based clustering algorithm DBSCAN to quantify data value factors and achieve differentiated scheduling. Design congestion awareness and hierarchical congestion control strategies. Build a scenario-adaptive network congestion index based on bandwidth utilization, latency jitter, and queue length change rate. Establish a three-level congestion control strategy: transmit data based on value in low-congestion conditions; limit traffic and prioritize high-value data in moderate-congestion conditions; and suspend low-value data transmission and discard expired data in high-congestion conditions. Information timeliness model and fair transmission scheduling: constructing a weighted peak information age (PAoI) indicator that integrates data value weights to evaluate the information timeliness of data of different values; combining the Jain fairness index to construct a multi-sensor PAoI fairness function that integrates data value weights to ensure transmission fairness; Establish a dynamic energy evolution model for deep-sea observation and detection equipment, establish a dynamic energy acquisition model based on the seasonal and diurnal variation characteristics of solar power supply, design an energy consumption model combining transmission and dormant energy consumption, and establish a long-term energy budget balance mechanism under battery energy constraints; Construct a multi-objective optimization problem that jointly optimizes weighted PAoI, energy consumption, and congestion; An optimization problem solving method based on DQN and imitation learning is adopted, and secure reinforcement learning is introduced to solve the optimal transmission strategy that meets the constraints through the Markov decision process.

2. The method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things according to claim 1, characterized in that: The multimodal data dynamic transmission model of deep-sea observation and detection equipment includes the following steps: In the multi-queue modeling and queue dynamic update, based on the data generation rules of ocean observation and detection equipment sensors, an independent data queue matching the number of sensors is established; periodic meteorological and hydrological observation data are generated at fixed time intervals to generate data packets. Obey uniform distribution; event-triggered target discovery, hydrological anomaly and meteorological warning data generate data packets according to the Poisson process, and the arrival volume The obedience parameter is Poisson distribution; build a multi-queue structure, and dynamically update the data queue length formula according to the data packet arrival and transmission process: ,in, For sensors At the moment The queue length, is the transmission decision variable, The amount of data transmitted in a single time, is the amount of newly arrived data, is the maximum queue length to avoid storage overflow; In the satellite narrowband communication bandwidth constraint modeling, according to the bandwidth limitation of the satellite narrowband communication module carried by the device, the upper limit of the total bandwidth resource in a single time slot is ,sensor The bandwidth requirement for a single transmission is , the narrowband communication bandwidth constraint formula is constructed as ; In constructing a dynamic data value weight model, the value weight factor of each sensor data is calculated based on the sensor data type, time characteristics and spatial characteristics; data value analysis based on spatiotemporal intelligence is adopted. In the time dimension, based on the LSTM network, the fluctuation amplitude of the marine environment data in the time dimension is perceived, abnormal changes in the data are identified, and the marine environment fluctuation rating is obtained. , the corresponding time weighting coefficient is In the spatial dimension, DBSCAN is used to cluster the spatial hotspots of high-precision navigation and positioning data, and regional division and pattern recognition of maritime space are performed to achieve spatial hotspot rating. , the corresponding spatial weighting coefficient is ; Through dynamic data value weighting factors , to achieve differentiated service quality assurance.

3. The method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things according to claim 1, characterized in that: The congestion perception and hierarchical congestion control strategy includes the following steps: In the congestion awareness model, real-time monitoring of satellite link bandwidth utilization , reflecting the current channel load; calculating the difference between the current round-trip time RTT and the minimum RTT, characterizing the intensity of network fluctuations , effectively identify the unstable state of the link, the maximum delay jitter threshold allowed by the network is ; Queue length change rate , the weighted average of the change rate of each sensor queue length can be obtained , characterizes the queue change trend; calculates the congestion index based on three indicators: bandwidth utilization, delay jitter, and queue length change rate , , dynamically adjust the weight according to the characteristics of the ocean environment, and increase Weight, data transmission peak period increases Weight, increased when sudden data influx weight; According to the real-time congestion level , design a hierarchical congestion control strategy; when When low congestion mode is used, the data value factor Determines data transmission priority and allows maximum transmission volume ;when , medium congestion model, start traffic shaping, limit the upper limit of single transmission data volume, and give priority to transmitting high-value data; when When the queue is close to overflow, the oldest data packet is discarded according to the timeliness; in the total time statistical period The long-term average congestion index of the system is .

4. The method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things according to claim 1, characterized in that: The information timeliness model and fair transmission scheduling include the following steps: In information timeliness modeling, information age PAoI is an indicator to measure information timeliness. No. The peak information age PAoI of a data packet is recorded as ,in, is the transmission completion time, The data generation time; there are Sensors, the data value weight factor is introduced into the system average PAoI calculation to form the weighted PAoI indicator formula: ,in, is the total number of packets transmitted; In the PAoI-driven multi-sensor data fairness transmission scheduling optimization method, multiple sensors have different data collection cycles, data volumes, and data values. In order to quantify timeliness fairness, the Jain fairness index is introduced and combined with the data value weight factor to construct a multi-sensor information timeliness fairness function. The formula is: .

5. The method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things according to claim 1, characterized in that: The energy dynamic evolution model of the deep-sea observation and detection equipment includes the following steps: In the energy acquisition model, based on the characteristics of solar power supply, Modeling the energy harvesting process for an exponential distribution of parameters , according to the day-night cycle Adjust the energy acquisition model, The length of daylight is dynamically adjusted with seasonal changes; In the energy consumption model, the sensor Single transmission consumption and sleep energy consumption satisfy , in the total time statistical period The long-term average energy consumption of the system is ; In battery energy constraints, define the upper and lower limits of battery capacity and , dynamically update the remaining energy ; Taking into account the battery life and usage stability of deep-sea observation and detection equipment, the energy fluctuation range is limited, and the equipment meets the long-term energy balance constraints.

6. The method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things according to claim 1, characterized in that: The multi-objective optimization problem of jointly optimizing weighted PAoI, energy consumption, and congestion includes the following steps: In constructing the joint optimization objective function, in order to achieve the optimal trade-off between system information timeliness, energy consumption and congestion, a multi-objective joint optimization problem is constructed. The objective function is to minimize the weighted sum of weighted PAoI, energy consumption and congestion. The formula is: ,in, is the energy consumption penalty coefficient, is the congestion penalty coefficient, which is used to adjust the balance between weighted PAoI, energy consumption and congestion. DQN and imitation learning are used to solve the optimization problem.

7. The method for scheduling multimodal data narrowband transmission of ocean satellite Internet of Things according to claim 1, characterized in that: The optimization problem solving method based on DQN and imitation learning includes the following steps: In the modeling of the interaction between the agent and the environment, the data transmission scheduling problem is modeled as a Markov decision process. The agent perceives the current state and performs actions by interacting with the environment. The state space includes the sensor queue length, the current congestion index, the current device energy state, the data value weight factor and the weighted PAoI. The action space includes the transmission decision variables And the amount of data transferred in a single time Combined with secure reinforcement learning, a security constraint layer is introduced to constrain actions, including bandwidth constraints and battery energy constraints. The reward function is the negative value of the joint optimization objective to minimize weighted PAoI, energy consumption, and congestion penalty. In the strategy of combining DQN with imitation learning, demonstration data of expert transmission scheduling decisions, including historical state sequences and action sequences, is collected to construct an expert dataset. The expert dataset is used to pre-train the DQN network through supervised learning to initialize the Q-value function parameters. During the DQN training process, the expert action probability distribution obtained by imitation learning is combined with samples in the experience replay mechanism to generate hybrid training data. Based on the current state and model convergence, the exploration rate parameter is dynamically adjusted to reduce the cost of ineffective exploration. Through the training model combining DQN with imitation learning, DQN can quickly learn the action distribution that meets the constraints and adapt to the complex dynamic ocean environment.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images