System and method for cooperative communication and predictive seamless switching of star network and cellular network

Abstract

The invention provides a star network and cellular network cooperative communication and predictive seamless switching system and method. The method comprises the steps that S1, a dynamic environment database is loaded, and a star network and cellular network communication module is self-checked; s2, continuously collecting state parameters of a star network communication link and a cellular network communication link; calculating a double-link quality prediction curve in a period of time in the future; s3, if the double-link quality prediction curve obtained through calculation meets a preset switching condition, generating a switching instruction; and S4, receiving a switching instruction, copying the data packet with the highest priority in the switching time window, sending the data packet on the two links at the same time, and switching the data packets with the other priorities to a target link. By converting a switching mechanism into active prediction from passive response, the method can avoid communication interruption caused by predictable events such as satellite shielding and entering a honeycomb blind area in advance, fundamentally guarantees continuity of formation control, and has extremely high communication reliability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of Internet of Vehicles and automatic driving communication, in particular to a system and method for cooperative communication and predictive seamless switching of satellite network and cellular network for vehicle platoon. BACKGROUND

[0002] Vehicle platoon driving is one of the key technologies of intelligent transportation system, which enables vehicles to share control information through inter-vehicle wireless communication (V2V) and to drive closely in small distance, thereby significantly improving road traffic efficiency, saving energy and improving driving safety. Currently, cellular Internet of Vehicles technology (C-V2X / 5G) is the mainstream solution to realize V2V communication of vehicle platoon, which has the advantages of low delay and high bandwidth. However, the ground cellular network has a fixed limitation in geographical coverage, and there are often signal coverage blind areas or unstable signal quality problems in areas such as highways, remote suburbs, mountainous areas or tunnels, resulting in communication interruption.

[0003] To make up for the insufficient coverage of the ground network, low-orbit satellite network is introduced as a supplementary communication means due to its wide-area coverage. However, simply combining satellite network and cellular network will introduce new technical challenges:

[0004] 1. Link interruption risk: Satellite network communication will experience a short communication interruption when switching to the next visible satellite or the signal is blocked by tall buildings, mountains, overpasses, etc. Similarly, vehicles will also experience connection interruption when entering or leaving the edge area of cellular network coverage. For vehicle platoon with high-speed driving and extremely close distance, any loss of critical control information may trigger a chain reaction, posing a serious safety hazard.

[0005] 2. Switching delay and data packet loss: Traditional communication switching mechanisms are mostly "first disconnected and then connected" or passive switching based on the current signal strength threshold. Such mechanisms have switching delay and data packet loss of tens to hundreds of milliseconds, which cannot meet the stringent requirements of ultra-low latency less than 20 milliseconds and zero packet loss for vehicle platoon cooperative control.

[0006] 3. Differences in link characteristics and resource waste: The delay, jitter, bandwidth, etc. of satellite network and cellular network are different and dynamically changing. If all data is transmitted on a single link without selection, not only the comprehensive advantages of dual links cannot be brought into play, but also the transmission quality of certain services (such as high-bandwidth perception data sharing and ultra-low-delay control instructions) may be affected due to improper link selection.

[0007] Therefore, existing technologies lack a collaborative communication mechanism that can intelligently anticipate future changes in communication link quality, plan ahead, and execute seamless switching to ensure ultimate safety and operational efficiency of vehicle platoons in complex and ever-changing communication environments.

[0008] Patent document CN116684886B discloses a cellular network planning method for a wargame board, which allows satellites to be displayed on the board through corresponding squares; it organizes the main satellite, ground station, and subordinate satellites into a satellite cellular network, ultimately making the visualization and simulation of the satellite cellular network on the board more intuitive. However, this patent document does not solve the aforementioned problem. Summary of the Invention

[0009] To address the shortcomings of existing technologies, the purpose of this invention is to provide a system and method for collaborative communication and predictive seamless handover between satellite networks and cellular networks.

[0010] A system for collaborative communication and predictive seamless handover between satellite network and cellular network, provided by the present invention, includes:

[0011] Hybrid communication and predictive handover unit;

[0012] The hybrid communication and predictive handover unit includes:

[0013] The link quality monitoring module is used to collect the physical layer and network layer performance parameters of star network communication links and cellular network communication links;

[0014] A predictive modeling and decision engine is used to store environmental data and combine it with performance parameters collected by the link quality monitoring module to obtain a comprehensive quality score sequence for star network communication links and cellular network communication links over a future period; and to generate decision instructions based on the comprehensive quality score sequence.

[0015] The intelligent data routing and execution module is used to receive and execute decision instructions.

[0016] Preferred options also include:

[0017] The data priority partitioning module is used to classify and label the processor's data;

[0018] C-V2X communication module, used for system connection with terrestrial cellular network;

[0019] StarNet terminal, used for connecting the system to low-Earth orbit satellite networks;

[0020] The processor is used to run formation control algorithms and control the movement of vehicles.

[0021] Preferably, the predictive modeling and decision engine includes:

[0022] A dynamic environment database is used to store environmental data, including real-time vehicle positioning, navigation paths, 3D map data, satellite ephemeris data, and historical network coverage quality data.

[0023] The link quality prediction model consists of one or a group of machine learning models. It takes real-time data collected by the link quality monitoring module and future environmental information in the dynamic environment database as input, and outputs a comprehensive quality score sequence of two links within the next N seconds.

[0024] The switching decision logic generates a decision instruction based on the comprehensive quality score sequence, combined with a preset switching threshold and a lag judgment mechanism.

[0025] Preferably, the decision instructions include preparing for switching, executing switching, and remaining unchanged; the switching decision logic also includes an emergency triggering mechanism to deal with sudden link quality degradation.

[0026] Preferably, the data priority classification module divides the processor's data into at least three priority categories: critical control, collaborative perception, and general business.

[0027] Preferably, the intelligent data routing and execution module is used to manage data flows, and its strategies include:

[0028] During the switching window, critical control data packets are copied and distributed simultaneously through two links to ensure redundant transmission of critical information.

[0029] When the dual links are stable, link aggregation and load balancing are performed based on the data characteristics and link characteristics of three priorities.

[0030] When a single link is available, a degraded operation strategy is implemented to ensure core business operations.

[0031] A method for collaborative communication and predictive seamless handover between satellite network and cellular network provided by the present invention includes:

[0032] Step S1: When the vehicle starts, load the dynamic environment database and perform a self-check of the satellite network and cellular network communication module;

[0033] Step S2: While the vehicle is in motion, continuously collect the status parameters of the StarNet communication link and the cellular network communication link;

[0034] Calculate the dual-link quality prediction curve for a future period based on the state parameters and information from the environmental database.

[0035] Step S3: If the calculated dual-link quality prediction curve meets the preset switching conditions, then a switching command is generated;

[0036] Step S4: Receive the handover instruction, copy the highest priority data packet within the handover time window and send it on both links simultaneously, and switch the remaining priority data packets to the target link.

[0037] Preferred options also include:

[0038] Step S5: When the quality of both links is higher than the stability threshold, enter aggregation mode:

[0039] According to the preset strategy, the highest priority data is sent to the link with the lowest latency, and the data with other priorities is sent to the link with the highest bandwidth.

[0040] Preferably, step S3 includes:

[0041] When switching from a superior link to a inferior link, or recovering from a inferior link to a superior link, different scoring thresholds are used to form a buffer zone;

[0042] When a sudden drop in the quality of the current link is detected, the prediction model is bypassed and a switchover is triggered immediately.

[0043] Preferably, the switching conditions include:

[0044] It is predicted that the current link will fall below a threshold in the next T seconds, while the target link will rise above another threshold.

[0045] Compared with the prior art, the present invention has the following beneficial effects:

[0046] 1. By changing the switching mechanism from "passive response" to "active prediction", this invention can avoid communication interruptions caused by foreseeable events such as satellite obstruction and entering cellular blind spots in advance, fundamentally ensuring the continuity of formation control and having extremely high communication reliability.

[0047] 2. This invention features a unique "copy-distribution" switching execution strategy, which ensures that the highest priority control data stream is delivered with 100% redundancy during the switching transition. This achieves a seamless switching experience with zero packet loss and zero interruption that is transparent to upper-layer applications.

[0048] 3. The intelligent routing and link aggregation capabilities in this invention enable the system to dynamically match the most suitable service to the most suitable link, making full use of the complementary advantages of StarNet and cellular networks. While ensuring security, it supports larger-scale data interaction, thereby allowing for more efficient and intelligent formation and coordination strategies, which can maximize system operating efficiency.

[0049] 4. Enhanced system robustness and security: The system of this invention has the ability to autonomously adapt to complex communication environments. Before entering a known weak signal area, it can automatically adjust the data transmission strategy and notify the upper layer application (such as appropriately increasing the safe distance between vehicles), thereby improving the safety redundancy and robustness of the entire formation system under extreme conditions. Attached Figure Description

[0050] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0051] Figure 1 This is a schematic diagram of the system architecture of the hybrid communication and predictive switching unit proposed in this invention.

[0052] Figure 2 This is a flowchart illustrating the predictive seamless switching method proposed in this invention.

[0053] Explanation of reference numerals in the attached figures

[0054] Hybrid communication and predictive switching unit 100 C-V2X communication module 105

[0055] Link quality monitoring module 101, Star Network terminal 106

[0056] Predictive Modeling and Decision Engine 102 Processor 107

[0057] Data Priority Division Module 103

[0058] Intelligent Data Routing and Execution Module 104 Detailed Implementation

[0059] The present invention will now be described in detail with reference to specific embodiments. These embodiments will help those skilled in the art to further understand the present invention, but do not limit the invention in any way. It should be noted that those skilled in the art can make several changes and improvements without departing from the concept of the present invention. These all fall within the scope of protection of the present invention.

[0060] The main objective of this invention is to overcome the shortcomings of existing technologies and provide a system and method for collaborative communication and predictive seamless handover between satellite networks and cellular networks in vehicle platooning. This system is capable of:

[0061] Active prediction: Based on vehicle travel routes, high-precision maps, satellite ephemeris and real-time network parameters, accurately predict the trend of link quality changes between satellite network and cellular network in the future.

[0062] Seamless handover: Based on the prediction results, proactive handover is initiated in advance, and a key data replication and distribution mechanism is adopted to achieve zero packet loss of control information during the handover process, achieving a "zero-interruption" handover effect that is imperceptible to upper-layer applications.

[0063] Intelligent routing: When both links are available, data streams are dynamically allocated based on the priority of data services (such as control commands and sensing data) and the real-time characteristics of the links (latency and bandwidth) to achieve optimal utilization of link resources and optimization of overall formation performance.

[0064] Enhanced reliability: Establish a multi-layered redundancy guarantee mechanism that includes predictive handover, emergency handover, and single-link degradation operation to ensure the robustness and security of the system under any foreseeable or sudden network changes.

[0065] Reference Figure 1 As shown, the present invention proposes a Hybrid Communication and Predictive Switching Unit 100 (HCPSU), which serves as a core component of an Onboard Unit (OBU) or a domain controller.

[0066] In terms of systems, this invention proposes a cooperative communication system, which includes:

[0067] Link quality monitoring module 101: Configured for real-time, high-frequency collection of key performance parameters of the physical layer and network layer of star network communication links and cellular network communication links.

[0068] Predictive Modeling and Decision Engine 102: The core of the system, which includes:

[0069] Dynamic Environment Database: Stores real-time vehicle location, navigation path, high-precision 3D map data, satellite ephemeris data, and historical network coverage quality data.

[0070] Link quality prediction model: One or a set of machine learning models (such as Long Short-Term Memory network LSTM) take real-time data collected by the monitoring module and future environmental information (such as obstructions on the future path, satellites to be switched) from the dynamic environment database as input, and output a sequence of comprehensive quality scores for the two links in the next N seconds [S(t+1), S(t+2), ... S(t+N)], where S = w·S latency +w·S loss +w·S jitter The weighting strategies differ for different data categories, such as critical control (L1), collaborative sensing (L2), and general business (L3). Machine learning models, such as LSTM, are used to predict latency, packet loss, and jitter scores within the next N seconds. Where E represents the vehicle's current GPS location, navigation waypoints in the next N seconds, 3D terrain / building occlusion data corresponding to the waypoints, the predicted time of satellite overpass handover, and the historical coverage strength of the C-V2X base station, etc.

[0071] Switching decision logic: Based on the predicted link quality score sequence, combined with preset switching thresholds and a hysteresis judgment mechanism, a decision instruction of "prepare for switching", "execute switching", or "remain unchanged" is generated. This logic also includes an emergency trigger mechanism to deal with sudden link quality degradation.

[0072] Data Priority Classification Module 103: Configured to classify and mark data packets to be sent generated by applications (such as formation control algorithms), and classify them into at least three priority categories: critical control category (L1), collaborative sensing category (L2), and general business category (L3).

[0073] Intelligent Data Routing and Execution Module 104: Configured to receive and execute instructions from the decision engine. This module is responsible for managing data flow, and its core strategies include:

[0074] During the switching window, L1 level data packets are copied and distributed simultaneously through both the old and new links to ensure redundant transmission of critical information.

[0075] When both links are stable, link aggregation and load balancing are performed based on the data characteristics of L1, L2, and L3 and the characteristics of the links.

[0076] When a single link is available, a degraded operation strategy is implemented to ensure core business operations.

[0077] like Figure 2 As shown, the present invention also proposes a cooperative communication method, which includes the following steps:

[0078] 1. Initialization: When the vehicle starts, the system loads the dynamic environment database and performs a self-check of the satellite network and cellular network communication modules.

[0079] 2. Continuous Monitoring and Prediction: During vehicle operation, the link quality monitoring module 101 continuously collects dual-link status parameters; the predictive modeling and decision engine 102, based on the collected parameters and dynamic environment database information, iteratively calculates and updates the dual-link quality prediction curve for the next N seconds. If the current time is t, and the historical scores of the previous M times and the environmental information of the next time are obtained, then the formula is used... Predict the link's latency, packet loss, and jitter scores for the next moment, and further utilize... Calculate the quality score for the next moment. By applying and repeating the above calculation process, the quality score of the two links in the next N seconds can be predicted, and a two-link quality prediction curve can be plotted.

[0080] 3. Switching Decisions:

[0081] Predictive triggering: The decision engine determines whether the predicted link quality curve meets the switching conditions (e.g., it is predicted that the current link will fall below a threshold in the next T seconds, while the target link will be above another threshold).

[0082] Lag judgment: To prevent frequent switching caused by fluctuations in link quality around the threshold, different scoring thresholds are used when deciding to switch from a superior link to a poor link and to recover from a poor link to a superior link, forming a stable buffer.

[0083] Emergency Trigger: If the monitoring module detects a sudden and sharp decline in the current link quality (such as a precipitous drop in RSRP value), it bypasses the prediction model and immediately triggers a switchover. The threshold used to determine a sudden and sharp decline in quality can be set using empirical values ​​from actual deployments.

[0084] 4. Seamless execution switching:

[0085] Upon receiving the switching instruction, the routing module copies the L1 level (critical control class) data packets and sends them simultaneously on both links within the preset switching time window.

[0086] L2 and L3 level data are directly switched from the original link to the target link.

[0087] After the handover window ends, data transmission on the original link stops, completing the handover. The receiving end performs deduplication processing on duplicate L1 data packets.

[0088] 5. Link Aggregation and Intelligent Routing: When the decision engine determines that the quality of both links is above the stable threshold, it enters aggregation mode. The routing module, based on a preset strategy, sends the most latency-sensitive L1 data to the link with the lowest latency (usually C-V2X), and sends the bandwidth-intensive L2 / L3 data to the link with the highest bandwidth (usually StarNet).

[0089] Example 1

[0090] Reference Figure 1 In this embodiment, the proposed Hybrid Communication and Predictive Switching Unit 100 (HCPSU) is installed in each member vehicle of the vehicle platoon. The HCPSU acquires its own vehicle status (speed, acceleration, steering wheel angle, etc.) via the vehicle's CAN bus and interacts with the platoon control algorithm on the application processor 107. The system connects to a low-Earth orbit satellite network via a satellite network terminal 106 and to a terrestrial cellular network via a C-V2X communication module 105.

[0091] The Hybrid Communication and Predictive Switching Unit 100 (HCPSU) mainly consists of the following four modules:

[0092] 1. Link quality monitoring module 101:

[0093] This module is a data acquisition unit that acquires and quantifies the link status from the underlying driver interface of the Star Network Terminal 106 and the C-V2X communication module 105 at a frequency of not less than 10Hz (i.e., once every 100ms).

[0094] StarNet link parameters include: Signal-to-Noise Ratio (SNR), Received Signal Strength Indication (RSSI), Bit Error Rate (BER), Round-Trip Delay (Ping RTT), Delay Jitter, Identifier (ID) of the currently serving satellite, and the real-time elevation and azimuth angles of the satellite relative to the vehicle.

[0095] C-V2X link parameters include: Reference Received Power (RSRP), Reference Received Quality (RSRQ), Signal-to-Interference-plus-Noise Ratio (SINR), Cell ID of the serving base station, Cell Traffic Load, and network mode (such as 5G-SA or NSA). All collected parameters are timestamped and transmitted to the predictive modeling and decision engine 102.

[0096] 2. Predictive Modeling and Decision Engine 102:

[0097] This engine is the "brain" of the system, responsible for prediction and decision-making, including:

[0098] Dynamic Environment Database: This database integrates multiple information sources.

[0099] Navigation and positioning information: The vehicle's precise location, speed, heading, and planned driving route for the next few minutes, obtained from the onboard GPS / IMU.

[0100] High-precision 3D map: Includes road topology, slope, curvature, and 3D models of objects that may obstruct satellite or ground signals, such as buildings, mountains, and overpasses on both sides of the road.

[0101] Satellite ephemeris data: real-time or near-real-time acquisition of precise orbital parameters of all low-Earth orbit satellites from the network, used to calculate the position and flight trajectory of visible satellites at any location at any future time.

[0102] Historical network database: During vehicle operation, the measured C-V2X network quality can be correlated with geographic location information to form a dynamically updated network coverage heat map, providing historical experience data for prediction.

[0103] Link quality prediction model: In this embodiment, an offline-trained Long Short-Term Memory (LSTM) network model is used.

[0104] Model Training: Supervised learning is performed using a large amount of historical data. The input (X) to the training data includes: the sequence of link quality parameters at historical moments, the corresponding vehicle positions, paths, and the satellite positions and map occlusion at that time. The label (Y) of the training data is: the link quality parameters actually measured within the next 1-5 seconds. Through training, the model learns the complex nonlinear relationship between various environmental factors and future link quality.

[0105] Model inference: During actual vehicle operation, the latest real-time data stream collected by the link quality monitoring module 101, as well as the vehicle's position (based on the navigation path) for the next 5 seconds extracted from the dynamic database, map occlusion information on the path, and the trajectory of the serving satellites and possible satellite switching events within the next 5 seconds are provided as input to the LSTM model.

[0106] Model Output: The model outputs two prediction curves. The comprehensive quality score sequences for the StarNet and C-V2X link within the next 5 seconds are as follows:

[0107] Score_Starnet(t+1s)...Score_Starnet(t+5s)

[0108] and

[0109] Score_C-V2X(t+1s)...Score_C-V2X(t+5s).

[0110] The score is a weighted average of predicted bandwidth, latency, and stability, and ranges from 0 to 100.

[0111] Switching decision logic:

[0112] Predictive triggering: The decision logic continuously analyzes and predicts the scoring sequence. For example, one triggering rule is: if Predicted_Score_Starnet(t+3s) < 40 and Predicted_Score_C-V2X(t+3s) > 60, then generate a "Prepare to switch to C-V2X" instruction. The choice of the switching timing (3 seconds later in this example) is to allow sufficient preparation time.

[0113] Lag Decision Mechanism: To avoid repeated switching effects, the thresholds for switching out and switching in are different. For example, the trigger condition for switching from StarNet to C-V2X is that the StarNet score is below 40 points, but if you want to switch back from C-V2X to StarNet, the StarNet score must recover to above 60 points. The interval [40, 60] in between constitutes a stable buffer zone.

[0114] Emergency Trigger: This is a redundancy safety mechanism. If Current_RSRP < -110dBm and Previous_RSRP > -95dBm (a sharp drop occurs), an "emergency switchover" instruction is immediately generated, bypassing the prediction waiting period.

[0115] 3. Data Priority Division Module 103:

[0116] This module categorizes the data from application processor 107.

[0117] L1 - Critical Control and Safety: Acceleration / deceleration commands within the formation, target steering angle, braking requests, cooperative lane-changing intentions, emergency braking warnings (BSW), etc. Requirements: Latency <20ms, reliability >99.999% (zero packet loss).

[0118] L2-Collaborative Sensing: Raw or target-level data from cameras, LiDAR, and millimeter-wave radar, compressed and feature-extracted, for environmental perception sharing. Requirements: High bandwidth, latency <100ms.

[0119] L3 - General Business Functions: Downloading differential update packages for high-precision maps, updating vehicle software, uploading diagnostic logs, etc. Requirements: Reliable transmission is sufficient; latency is not a concern.

[0120] 4. Intelligent Data Routing and Execution Module 104:

[0121] This module is the executor of decisions.

[0122] The "Make-Before-Break" strategy: Upon receiving a "Switchover" instruction, this module initiates a brief (e.g., 200ms) switching window. Within this window, all L1 packets are duplicated and simultaneously transmitted via both the old link that is about to be disconnected and the newly established link. The receiving end's HCPSU is responsible for identifying and discarding duplicate packets. This approach trades brief bandwidth redundancy for absolute continuity of critical data flows.

[0123] "Link Aggregation and Load Balancing" Strategy: When the decision engine determines that both links are stable (e.g., both scores > 70), the module enters aggregation mode. According to the preset strategy, L1 packets are routed to the link with the lowest measured latency, while high-volume L2 and L3 packets are routed to the link with the highest available bandwidth, thus optimizing resources.

[0124] The "single-link degraded operation" strategy: When it is predicted that only one link will be available for a relatively long period of time (e.g., more than 10 seconds), the module will notify the application processor 107 to enter degraded operation mode. For example, it will temporarily suspend the sharing of L2 level perception data, retain only the transmission of L1 level core control information, and suggest that the formation control algorithm adopt a more conservative strategy, such as increasing the safe distance between vehicles.

[0125] Example 2

[0126] Reference Figure 2 The workflow of the method of the present invention is as follows:

[0127] Step S201: System initialization and self-test.

[0128] Step S202: Execute repeatedly. The link quality monitoring module 101 continuously collects real-time parameters of the two links.

[0129] Step S203: The predictive modeling and decision engine 102 generates a dual-link quality prediction score curve for the next N seconds based on real-time parameters and a dynamic environment database.

[0130] Step S204: The decision logic determines whether the switching conditions are met.

[0131] First, check if the emergency handover condition (S204a) has been triggered. If so, proceed directly to step S206 to perform the handover.

[0132] If not, check if the predictive switching condition (S204b) is met, and apply the hysteresis judgment logic. If yes, generate a switching instruction and jump to step S206.

[0133] Step S205: If no switching conditions are met, determine whether the link aggregation conditions are met (both links are excellent). If yes, enter intelligent routing mode (S205a) and distribute data according to data priority and link characteristics. If not, maintain the current link policy. Then return to step S202.

[0134] Step S206: Perform seamless handover. The intelligent data routing and execution module 104 initiates the handover window and replicates and distributes the L1 data.

[0135] Step S207: After the switch is completed, the system continues to operate with the new link configuration and returns to step S202.

[0136] Through the above system architecture and working method, this invention constructs a closed-loop "perception-prediction-decision-execution" intelligent communication management system, which effectively solves the core pain points faced by vehicle platooning in a hybrid network environment, such as communication interruption, switching delay and data packet loss, and provides a solid communication foundation for achieving safe and efficient autonomous driving platooning.

[0137] Those skilled in the art will understand that, besides implementing the system and its various devices, modules, and units provided by this invention in the form of purely computer-readable program code, the same functions can be achieved entirely through logical programming of the method steps, making the system and its various devices, modules, and units of this invention function in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers. Therefore, the system and its various devices, modules, and units provided by this invention can be considered as a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; alternatively, the devices, modules, and units for implementing various functions can be considered as both software modules implementing the method and structures within the hardware component.

[0138] Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. Unless otherwise specified, the embodiments and features described in this application can be arbitrarily combined with each other.

Claims

1. A system for collaborative communication and predictive seamless handover between satellite network and cellular network, characterized in that, include: Hybrid communication and predictive handover unit (100); The hybrid communication and predictive handover unit (100) includes: The link quality monitoring module (101) is used to collect the performance parameters of the physical layer and network layer of the StarNet communication link and the cellular network communication link; A predictive modeling and decision engine (102) is used to store environmental data and, in conjunction with performance parameters collected by the link quality monitoring module (101), obtain a comprehensive quality score sequence for the star network communication link and the cellular network communication link in the future; and generate decision instructions based on the comprehensive quality score sequence. The intelligent data routing and execution module (104) is used to receive and execute decision instructions.

2. The system for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 1, characterized in that, Also includes: The data priority partitioning module (103) is used to classify and label the data of the processor (107); C-V2X communication module (105) is used for the connection between the system and the terrestrial cellular network; StarNet Terminal (106) is used for connecting the system to a low-Earth orbit satellite network; The processor (107) is used to run the formation control algorithm and control the movement of the vehicles.

3. The system for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 1, characterized in that, The predictive modeling and decision engine (102) includes: A dynamic environment database is used to store environmental data, including real-time vehicle positioning, navigation paths, 3D map data, satellite ephemeris data, and historical network coverage quality data. The link quality prediction model consists of one or a group of machine learning models. It takes real-time data collected by the link quality monitoring module (101) and future environmental information in the dynamic environment database as input, and outputs a comprehensive quality score sequence of two links within the next N seconds. The switching decision logic generates a decision instruction based on the comprehensive quality score sequence, combined with a preset switching threshold and a lag judgment mechanism.

4. The system for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 3, characterized in that, The decision instructions include preparing for switching, executing the switching, and remaining unchanged; the switching decision logic also includes an emergency triggering mechanism to deal with sudden link quality degradation.

5. The system for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 2, characterized in that, The data priority division module (103) divides the data of the processor (107) into at least three priority categories: key control category, collaborative perception category, and ordinary business category.

6. The system for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 5, characterized in that, The intelligent data routing and execution module (104) is used to manage data flows, and its strategies include: During the switching window, critical control data packets are copied and distributed simultaneously through two links to ensure redundant transmission of critical information. When the dual links are stable, link aggregation and load balancing are performed based on the data characteristics and link characteristics of three priorities. When a single link is available, a degraded operation strategy is implemented to ensure core business operations.

7. A method for collaborative communication and predictive seamless handover between satellite network and cellular network, characterized in that, include: Step S1: When the vehicle starts, load the dynamic environment database and perform a self-check of the satellite network and cellular network communication module; Step S2: While the vehicle is in motion, continuously collect the status parameters of the StarNet communication link and the cellular network communication link; Calculate the dual-link quality prediction curve for a future period based on the state parameters and information from the environmental database. Step S3: If the calculated dual-link quality prediction curve meets the preset switching conditions, then a switching command is generated; Step S4: Receive the handover instruction, copy the highest priority data packet within the handover time window and send it on both links simultaneously, and switch the remaining priority data packets to the target link.

8. The method for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 7, characterized in that, Also includes: Step S5: When the quality of both links is higher than the stability threshold, enter aggregation mode: According to the preset strategy, the highest priority data is sent to the link with the lowest latency, and the data with other priorities is sent to the link with the highest bandwidth.

9. The method for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 7, characterized in that, Step S3 includes: When switching from a superior link to a inferior link, or recovering from a inferior link to a superior link, different scoring thresholds are used to form a buffer zone; When a sudden drop in the quality of the current link is detected, the prediction model is bypassed and a switchover is triggered immediately.

10. The method for collaborative communication and predictive seamless handover between satellite network and cellular network according to claim 7, characterized in that, The switching conditions include: It is predicted that the current link will fall below a threshold in the next T seconds, while the target link will rise above another threshold.

Images