Multi-communication link management platform and method for wireless ad hoc network

Abstract

The invention discloses a multi-communication link management platform and method for a wireless ad hoc network, and relates to the technical field of wireless communication, and the platform comprises a situation awareness module which is used for collecting data and generating situation data in combination with a three-dimensional digital map; the link occlusion risk prediction module is used for generating a predictive occlusion index based on the situation data; the multi-link comprehensive efficiency evaluation module is used for receiving the predictive shielding index and calculating and outputting a link comprehensive efficiency score in combination with the operation cost data; and the predictive switching decision and execution module is used for responding to the switching condition and switching the service data flow from the currently activated link to the inactivated link with the highest link comprehensive performance score. According to the method, communication interruption is avoided by actively predicting switching, expensive links are used as required, a multi-target model is adopted, the cost is reduced, and the method can flexibly adapt to service requirements.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a multi-communication link management platform and method of wireless ad hoc network. BACKGROUND

[0002] In large open pit mine area, automated port and other scenarios, unmanned vehicle cluster collaborative operation relies on stable and low-latency wireless communication network. However, in a complex electromagnetic environment, frequent movement of large metal equipment will cause instantaneous and deep signal fading and shielding of communication link, so there is a high requirement for communication reliability.

[0003] The existing multi-link management technology mostly adopts passive switching strategy based on received signal strength indication, which can only switch after communication quality degradation or even interruption, which is easy to cause key signaling loss and affect operation efficiency. However, if the expensive link is activated for a long time as a hot backup to make up for the hysteresis, although the communication reliability will be higher, it will also cause high traffic cost and occupation of public network resources, etc. The existing technology cannot balance the contradiction between communication reliability and operating cost, and a link management scheme with foresight and economy is urgently needed. SUMMARY

[0004] The purpose of the present application is to provide a multi-communication link management platform and method of wireless ad hoc network, which solves the problems in the background art.

[0005] To solve the above technical problems, the present application provides a multi-communication link management method of wireless ad hoc network, comprising the following steps: S1, for the unmanned vehicle node in the network, collecting the real-time three-dimensional position coordinates, velocity vector and predetermined travel trajectory of the unmanned vehicle node, and obtaining the real-time position and velocity of other unmanned vehicle nodes in the network as dynamic obstacle data, combining with the pre-set three-dimensional digital map, generating situation data; S2, based on the situation data, quantitatively predicting the risk of being shielded in a preset future prediction time window for the available communication link of the unmanned vehicle node, and generating a predictive shielding index; S3, receiving the predictive shielding index, and combining the real-time service quality parameters obtained from the network protocol stack and the preset operating cost data, calculating and outputting the link comprehensive performance score for the available communication link; S4, in response to meeting the preset switching condition, switching the business data flow from the currently activated link to the non-activated link with the highest link comprehensive performance score; The preset switching condition is that the non-activated link with the highest link comprehensive performance score is still higher than the sum of the link comprehensive performance score of the currently activated link and a preset switching delay threshold after a preset stable time.

[0006] Preferably, the step of generating the predictive blocking index comprises: S21, calculating a line-of-sight blocking component representing the possibility of a line-of-sight path being physically blocked based on the situation data; S22, calculating a diffraction influence component representing the proximity of the line-of-sight path to the dynamic obstacle based on the situation data; S23, combining the line-of-sight blocking component and the diffraction influence component according to preset risk model weights to generate the predictive blocking index.

[0007] Preferably, the step of calculating the link comprehensive performance score comprises: S31, converting the predictive blocking index generated in step S2 into a risk avoidance score; S32, calculating a normalized service quality score based on the real-time service quality parameters; S33, calculating a normalized inverse cost score based on the preset operating cost data; S34, combining the risk avoidance score, the normalized service quality score, and the normalized inverse cost score according to preset policy weights to generate the link comprehensive performance score.

[0008] Preferably, the step of calculating the normalized service quality score comprises: S321, combining the link latency and jitter in the real-time service quality parameters according to preset weights to calculate an original service quality score; S322, normalizing the current original service quality score using the observed historical maximum and minimum original service quality scores within a preset time period to generate the normalized service quality score; the length of the preset time period should ensure that service quality extrema under different network loads and communication environments can be collected to provide effective reference benchmarks for normalization.

[0009] Preferably, the step of calculating the normalized inverse cost score comprises: S331, retrieving the cost value corresponding to the current link from a preset cost table; the cost table can be dynamically obtained and updated through operator tariff standards, historical traffic statistics, or real-time billing interfaces; S332, normalizing the retrieved cost value using the highest cost value among all available links to generate the normalized inverse cost score.

[0010] Preferably, the preset policy weight is configurable and adjustable between a cost-sensitive mode, a safety-first mode and a performance-first mode according to the operation business target.

[0011] Preferably, the duration of the future prediction time window is set to be greater than the total time required for the system to perform a complete link switching process, to ensure a reserved safety margin for switching before the predicted link interruption actually occurs.

[0012] Preferably, the diffraction impact component is calculated based on the shortest distance between the line-of-sight path and all dynamic obstacles, and is determined with reference to a preset safety distance threshold value; the safety distance threshold value is determined according to the Fresnel zone radius corresponding to the operating frequency of the communication link.

[0013] Also provided is a multi-communication link management platform for a wireless ad hoc network, comprising: a situation awareness module, configured to collect real-time three-dimensional position coordinates, velocity vectors and predetermined travel trajectories of unmanned vehicle nodes in the network, and obtain real-time positions and velocities of other unmanned vehicle nodes in the network as dynamic obstacle data, and generate situation data in combination with a preloaded three-dimensional digital map; a link occlusion risk prediction module, configured to quantitatively predict the occlusion risk of the available communication links of the unmanned vehicle nodes within a preset future prediction time window based on the situation data generated by the situation awareness module, and generate a predictive occlusion index; a multi-link comprehensive performance evaluation module, configured to receive the predictive occlusion index generated by the link occlusion risk prediction module, and calculate and output the link comprehensive performance score for the available communication links in combination with real-time service quality parameters obtained from a network protocol stack and preset operation cost data; a predictive switching decision and execution module, configured to switch the traffic data flow from a currently activated link to a non-activated link with the highest link comprehensive performance score in response to the link comprehensive performance score meeting a preset switching condition; the preset switching condition is that the non-activated link with the highest link comprehensive performance score has a link comprehensive performance score that is still higher than the sum of the link comprehensive performance score of the currently activated link and a preset switching hysteresis threshold value after a preset stable time.

[0014] Compared with the prior art, the present application has the following beneficial effects: 1. Active management of communication links is achieved through spatiotemporal situation prediction, changing from passive to active, completing switching before the link quality decreases, actively avoiding interruption, guaranteeing stable cooperative control signaling delay, and improving the efficiency and safety of unmanned cluster cooperative operation.

[0015] 2. Realize the accurate short-time use of cellular network resources on demand, only enable the expensive link when the low-cost link will fail, release after the risk is removed, compared with the long-term "hot backup" scheme, can effectively avoid the waste of network resources and reduce the operating cost, realize the optimization control of operating cost.

[0016] 3. Adopt multi-objective evaluation model, unify multiple dimensions such as risk, performance, cost in one framework for quantitative evaluation and decision, and allow the operator to flexibly adapt to different business needs by adjusting the strategy weight, with high flexibility and scalability. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, below will briefly introduce the drawings needed to be used in the embodiments or prior art description, obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the premise of these drawings; Figure 1 The flow chart of the method of the present application. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present application will be described clearly and completely below in combination with the drawings in the embodiments of the present application, obviously, the described embodiments are only some embodiments of the present application, not all the embodiments, based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0019] Embodiment 1 Please refer to Figure 1 The present application provides a multi-communication link management method of wireless ad hoc network, comprising the following steps: S1, for any unmanned vehicle node in the network, collecting the real-time three-dimensional position coordinates, velocity vector and predetermined driving trajectory of the unmanned vehicle node, and obtaining the real-time position and velocity of other unmanned vehicle nodes in the network as dynamic obstacle data, combining with the pre-set three-dimensional digital map, generating situation data; S2, based on the situation data, for each available communication link of the unmanned vehicle node, quantitatively predicting the risk of being blocked in the preset future prediction time window, generating a predictive blocking index; S3, receiving the predictive blocking index, and combining the real-time service quality parameters obtained from the network protocol stack and the preset operating cost data, calculating and outputting the link comprehensive performance score for each available communication link; S4, in response to the preset switching condition being met, switching the service data stream from the currently activated link to the non-activated link with the highest link comprehensive performance score; The preset switching condition is that the non-activated link with the highest link comprehensive performance score has a link comprehensive performance score that is still higher than the sum of the link comprehensive performance score of the currently activated link and a preset switching hysteresis threshold after lasting a preset stabilization time. The preset stabilization time is determined comprehensively according to the following factors: vehicle moving speed, obstacle density, link switching execution time length, and service tolerance to switching frequency, and specifically, a time threshold that can filter out transient fluctuations while ensuring timely response is selected by statistically analyzing the link stability duration distribution in historical switching data.

[0020] The preset switching hysteresis threshold is determined by analyzing the fluctuation characteristics of the link comprehensive performance score, specifically, the fluctuation amplitude of the performance score of different links in the same scene is collected, the standard deviation is calculated, and a difference threshold that can avoid ping-pong switching and will not delay necessary switching is set by combining the switching overhead and the service continuity requirement.

[0021] The method provided by the embodiment is a multi-communication link management method of a wireless ad hoc network, and aims to solve the technical contradiction that the unmanned vehicle cluster fails to complete a task or has high operating cost due to untimely communication switching in a dynamic occlusion environment. The core of the method is a forward-looking decision mechanism. First, the spatiotemporal situation awareness in step S1 lays a data foundation for subsequent accurate prediction. The unmanned vehicle node not only perceives its own position, speed, and trajectory, but also captures dynamic information of surrounding collaborative vehicles through vehicle-to-vehicle communication, and fuses high-precision three-dimensional map data, so as to construct a complete and dynamic operation scene situation map, i.e., situation data. Then, instead of passively waiting for signal fading, the method actively predicts the future in step S2, quantifies the risk of each communication link being physically occluded in a very short time window in the future, and forms a key predictive occlusion index. Then, step S3 no longer relies on a single signal strength indicator, but fuses the predicted risk, real-time network service quality, and preset operating cost into a multi-objective link comprehensive performance score, achieving comprehensive quantification of the "goodness" of the link. Finally, in step S4, the system switches the service to the optimal backup link in advance and smoothly based on the performance score and a precise switching logic with hysteresis and stabilization time, before predicting that the current link will deteriorate. This strategy of changing from passive to active ensures that the end-to-end delay of collaborative control signaling can be stabilized within a safety threshold in complex scenes such as large mines or automated ports, thereby fundamentally improving the operation efficiency and safety of the unmanned driving cluster.

[0022] Embodiment 2 The step of generating the predictive occlusion index comprises: S21, calculating a line-of-sight occlusion component representing the possibility of physical obstruction of the line-of-sight path based on the situational data; S22, calculating a diffraction impact component representing the proximity of the line-of-sight path to dynamic obstacles based on the situational data; S23, combining the line-of-sight occlusion component and the diffraction impact component according to preset risk model weights to generate a predictive occlusion index; The step of calculating the link comprehensive performance score comprises: S31, converting the predictive occlusion index generated in step S2 into a risk avoidance score; S32, calculating a normalized service quality score based on real-time service quality parameters; S33, calculating a normalized inverse cost score based on preset operating cost data; S34, combining the risk avoidance score, the normalized service quality score and the normalized inverse cost score according to preset strategy weights to generate the link comprehensive performance score; The step of calculating the normalized service quality score comprises: S321, combining the link latency and the jitter in the real-time service quality parameters according to preset weights to calculate an original service quality score; S322, normalizing the current original service quality score using the observed historical maximum and minimum original service quality scores within a preset time period to generate the normalized service quality score, the length of the preset time period should ensure that the service quality extrema under different network loads and communication environments can be collected to provide effective reference benchmarks for the normalization; the time period should be long enough to cover the typical range of service quality changes and short enough to maintain sensitivity to the current network state; The preset time period is determined according to the following principles: analyze the characteristics of the vehicle operation cycle and the variation law of network load, select the minimum time span that can cover typical working condition changes to ensure that the normalization benchmark is both statistically representative and adaptable to the dynamic changes of network state.

[0023] The step of calculating the normalized inverse cost score comprises: S331, retrieving the cost value corresponding to the current link from a preset cost table; the cost table can be dynamically obtained and updated through the operator's tariff standard, historical traffic statistics or real-time billing interface; S332, normalizing the retrieved cost value using the highest cost value among all available links to generate the normalized inverse cost score; The available links include all link types that the unmanned vehicle node can currently establish a communication connection, such as the ad hoc network link (cost is 0) and the 5G cellular network link (charged by traffic) mentioned in the embodiments of the specification, wherein the highest cost value is the unit cost of the most expensive link (such as the 5G link).

[0024] The preset strategy weight can be configured and adjusted between the cost-sensitive mode, the security-first mode and the performance-first mode according to the operation business target; The duration of the future prediction time window is set to be greater than the total time required for the system to perform a complete link switching process, to ensure that a safety margin is reserved for switching before the predicted link interruption actually occurs; The embodiment sets forth the setting logic of the future prediction time window , which is a key constraint to ensure that the method is truly proactive; the duration of the time window is not arbitrarily set, but must be greater than the total time required for the system to perform a complete link switching process ; is a concept known in the art, which includes all the time required for signaling interaction, routing table update and a series of operations; by setting to a value greater than , for example, setting , the system ensures that there is sufficient safety margin; it is like a vehicle needs to start decelerating a distance before reaching an intersection, similarly, the method starts the decision and switching process in advance before predicting that the link will be "interrupted" at this "intersection"; for example, if a complete link switching takes 200 milliseconds, setting to 300 milliseconds means that the system can always complete the switching action at least 100 milliseconds before the link quality really collapses, thereby effectively avoiding packet loss and service interruption due to untimely switching; this reserved safety margin is the core guarantee to achieve "active avoidance" of interruption.

[0025] Embodiment 3 The diffraction influence component is calculated based on the shortest distance between the line-of-sight path and all dynamic obstacles, and is determined with reference to a preset safety distance threshold; the safety distance threshold is determined according to the Fresnel zone radius corresponding to the working frequency of the communication link; This embodiment illustrates the specific process of generating the predictive blockage index, which is the core of realizing proactive link switching; this process abandons the traditional way of relying on historical signal strength for judgment, and instead constructs a forward-looking physical risk assessment model; the motivation is that the motion trajectory of a physical entity is predictable, so the communication blockage it causes can also be predicted; the existing technology cannot predict the sudden communication interruption caused by the movement of physical entities, and the motivation of the present invention is to create a forward-looking risk assessment model; this model predicts the blocking effect of the link in the communication world by modeling the motion trajectory of the vehicle in the physical world; The generation of the predictive blockage index PBI is not a single-dimensional consideration, but a delicate integration of two key physical effects: one is the line-of-sight blockage component calculated by step S21, which represents the most serious "hard blockage" situation, i.e. the possibility of the line-of-sight path of communication being completely blocked by an obstacle; the second is the diffraction influence component calculated by step S22, which quantifies the signal diffraction attenuation caused by the obstacle being too close even if the line-of-sight path is not completely blocked, i.e. the risk of "soft blockage"; finally in step S23, the two components are combined into a unified risk measurement value through weighted combination; The predictive blockage index PBI is calculated by the following formula: ; represents the predictive blockage index of a specific link in a future prediction time window ; and represents the preset risk model weight, and .

[0026] Both of them determine the emphasis of the model on the following two kinds of blockage risks: "hard blockage" refers to the risk of the line-of-sight path of communication being completely blocked by a physical obstacle; "soft blockage" refers to the risk of significant diffraction effect caused by the proximity of the obstacle to the communication path; represents the line-of-sight blockage component, whose value is 0 or 1, 1 represents that the line-of-sight path is geometrically intersected by any obstacle at the predicted time, and 0 represents that it is not intersected; represents the shortest Euclidean distance between the line-of-sight path of communication and the surface of all obstacles in the prediction time window, which reflects the degree of diffraction influence; represents the preset safety distance threshold value for judging whether the obstacle constitutes a threat; the weight and The settings depend on specific business needs. In critical tasks where communication interruptions must be absolutely avoided, such as when unmanned mining truck convoys are operating in coordinated formation, settings can be configured. The weights are relatively high, such as 0.8, to prioritize avoiding the most dangerous hard occlusions. These weights can be optimized through offline training in a mining area digital twin simulation platform to minimize the prediction error rate. safe distance threshold The determination of this value is closely related to the Fresnel zone radius of radio waves and is adjusted based on practical engineering experience. It clearly defines how close an obstacle must be to the communication line-of-sight path to be considered a threat. The logic for setting this value is based on the standard calculation formula for the first Fresnel zone corresponding to the communication link's operating frequency, combined with engineering margins; for example, for 5.9 GHz DSRC communication, it can be... Set to 2 meters; This formula is periodically called by the link occlusion risk prediction module, for example, every 50 milliseconds, to calculate the PBI score for each candidate link l in the system; a PBI value close to 1 indicates that the link is within the prediction time window. It is highly likely that the link will be completely blocked; conversely, if the PBI value approaches 0, it means that the link will remain open within the window. Through this model, the system can quantify the complex physical blockage problem into a risk value between 0 and 1. This prior knowledge enables the system to clearly predict the future availability of the link, providing a basis for decision-making to complete the switch before the mining card enters the signal blind zone, thereby ensuring the reliable issuance of key commands such as emergency braking. The embodiment describes the calculation steps of the link comprehensive performance score, and the basic purpose of the steps is that a single risk prediction or a single performance indicator is not enough to make the optimal link switching decision; therefore, the system creates a unified, multi-objective utility function, which integrates the three mutually contradictory attributes of risk, performance, and cost into a directly comparable scalar value; the process first converts the aforementioned predictive blocking index into a risk aversion score through step S31, reflecting the consideration of future stability; then, step S32 captures the quality of service parameters from the network protocol stack in real time and calculates the normalized quality of service score, reflecting the current performance of the link; at the same time, step S33 calculates the normalized inverse cost score according to the preset cost table, and the operating economy is also included in the evaluation system; finally, step S34 generates the final link comprehensive performance score LUS according to the dynamically adjustable strategy weight, and the three scores are weighted and summed to generate the final link comprehensive performance score LUS; the essence of this design is that it allows the operator to switch to a safety priority mode when performing key collaborative tasks, or to switch to a cost-sensitive mode in regular transportation tasks according to specific business objectives, so that the link management strategy can flexibly adapt to different operation requirements and operating strategies, achieving high flexibility and intelligence; The model aims to create a unified, multi-objective utility function that skillfully integrates these mutually contradictory attributes, namely risk, performance, and cost, into a scalar value that can be directly compared, thereby providing a unique and comprehensive basis for decision-making; The link comprehensive performance score LUS(l) is defined by the following formula: ; represents the comprehensive performance score of the link ; represents the normalized quality of service score; represents the risk aversion score, and the higher the PBI value, the lower the score; is the normalized inverse cost score; represents the preset strategy weight, and the sum of the three is 1, which is the importance of performance, risk, and cost; Strategy weight is the core strategy configuration parameter of the system, which is set by the operator according to business objectives; for example, in the cost-sensitive mode, the weight of can be increased to make the system tend to use free ad hoc network links; in the safety priority mode, the weight of is increased to ensure the reliability of communication at any cost; in the performance priority mode, the weight of The weights are assigned to prioritize low latency and low jitter; these weights can be dynamically adjusted so that the entire link management system can accurately adapt to different tasks and operational strategies. This model uses a multi-link comprehensive performance evaluation module to calculate the LUS score of all available links in real time, providing direct and quantitative input for the subsequent handover decision module. Its application effect is revolutionary, ensuring that the system doesn't just select a link with "good signal," but rather the link that is optimal overall over a period of time. For example, when an unmanned mining truck convoy needs to traverse an area with potential signal obstruction, the system can be configured to a safety-first mode, thus improving signal strength. With a weight of 0.7, even if a high-bandwidth 5G link is more expensive, as long as its predicted risk is much lower than that of the soon-to-be-blocked self-organizing network link, the system will still switch decisively to ensure that the fleet will not disband due to communication interruption or cause a chain reaction of safety braking and operation stoppage due to failure of coordinated control. The intelligence and foresight of this decision-making are unmatched by traditional technologies. This embodiment further refines two key components of the Link Overall Performance (LUS) score: the Normalized Quality of Service (SMS) score. and normalized inverse cost fraction The calculation method ensures that raw data with different physical units and dimensions can be converted to a uniform scale for fair comparison. For the service quality score, step S321 first combines the raw latency and jitter parameters provided by the network protocol stack into a raw service quality score through preset weights α and β. This design allows the system to prioritize tasks based on the varying sensitivities of different services to latency or jitter. Next, step S322 employs a max-min normalization method, utilizing historical extreme values ​​observed over a period of time to map the original score to a range of 0 to 1, generating... This eliminates the dimensional differences in performance parameters between different link types. For the cost score, the calculation process is more direct: step S331 retrieves the cost value of the current link from a preset cost table; for example, the cost of a self-organizing network link is 0, while a 5G link is charged based on traffic. Then, step S332 uses the highest cost value among all available links as a benchmark to normalize the current cost, generating an inverse cost score. Thus, free links The value is 1, while the most expensive link is Approaching 0; this series of standardized processes is the technical prerequisite for achieving multi-dimensional data fusion for scientific decision-making. It transforms raw, heterogeneous data into calculable and comparable performance indicators. Original service quality score Normalized service quality score and normalized inverse cost fraction The calculation formula is as follows: ; ; ; and respectively represent the real-time delay and jitter of the link , which have been normalized before calculation; and represent preset weights, reflecting the importance of delay and jitter; represent the calculated original quality of service score; and represent the historical maximum and minimum values observed by the system in the past period of time; represent the cost value of the link in the preset cost table; represent the highest cost value in all available links.

[0027] Embodiment 4 A multi-communication link management platform of a wireless ad hoc network comprises: A situation awareness module is configured to collect real-time three-dimensional position coordinates, velocity vectors and predetermined travel trajectories of any unmanned vehicle node in the network, and obtain real-time positions and velocities of other unmanned vehicle nodes in the network as dynamic obstacle data, and generate situation data in combination with a preset three-dimensional digital map; this module is the data basis of the platform, and is responsible for real-time collection and fusion of multi-dimensional data of the vehicle, collaborative vehicles and three-dimensional environment, and provides high-quality original situation data for the entire system; A link occlusion risk prediction module is configured to, based on the situation data generated by the situation awareness module, quantitatively predict the occlusion risk of each available communication link of the unmanned vehicle node within a preset future prediction time window, and generate a predictive occlusion index; this module is the core analysis component of the platform, which uses the aforementioned PBI model to perform in-depth analysis on the situation data, and prospectively calculates the future occlusion risk of each link; A multi-link comprehensive performance evaluation module is configured to receive the predictive occlusion index generated by the link occlusion risk prediction module, and combine real-time quality of service parameters obtained from a network protocol stack and preset operation cost data to calculate and output a link comprehensive performance score for each available communication link; this module performs calculation through the LUS model, and outputs a comprehensive quantitative performance score, which is the final judgment of the “goodness” of the link; The predictive handover decision and execution module is configured to switch the service data flow from the currently activated link to the non-activated link with the highest link comprehensive performance score in response to the link comprehensive performance score meeting the preset handover condition; this module is an execution component of the platform, which continuously monitors the LUS score of all links and makes a judgment based on a precise logic containing a stabilization time and an advantage boundary, and only when the advantage of a new link is obvious enough and stable enough, the handover is triggered; The embodiment provides a wireless self-organizing network multi-communication link management platform for implementing the above method; the platform is composed of four closely coupled and cooperatively working core modules, forming a complete data processing and decision execution closed loop; the modular design divides the complex link management task into four clear steps of sensing, predicting, evaluating and executing, the responsibilities of the modules are clear, the data flow is clear, and a fundamental change from passive response to active prediction is realized, so that the traffic cost and channel occupancy rate of the cellular network can be reduced by at least 70% compared with the prior art, and the cooperative operation efficiency and safety of the unmanned vehicle cluster are significantly improved.

[0028] The above is only a preferred embodiment of the present application, and does not limit the present application in other forms. Any skilled person in the art can use the disclosed technical content to make changes or modifications to equivalent embodiments applied to other fields, but any simple modification, equivalent change and modification made to the above embodiment without departing from the technical solution content of the present application, according to the technical essence of the present application, still belongs to the protection scope of the technical solution of the present application.

Claims

1. A method for managing multiple communication links in a wireless ad hoc network, characterized in that, Includes the following steps: S1. For unmanned vehicle nodes in the network, collect the real-time three-dimensional position coordinates, velocity vectors and predetermined driving trajectories of the unmanned vehicle nodes, and obtain the real-time position and velocity of other unmanned vehicle nodes in the network as dynamic obstacle data. Combined with a preset three-dimensional digital map, generate situational data. S2. Based on the situational data, the available communication links of the unmanned vehicle node are quantitatively predicted to be blocked within a preset future prediction time window, and a predictive blocking index is generated. S3. Receive the predictive occlusion index, and combine it with the real-time service quality parameters obtained from the network protocol stack and the preset operating cost data to calculate and output the comprehensive link performance score for the available communication link. S4. In response to meeting the preset switching conditions, the service data flow is switched from the currently active link to the inactive link with the highest comprehensive performance score of the link. The preset switching condition is: the inactive link with the highest overall link performance score, after a preset stable period of time, still has an overall link performance score higher than the sum of the overall link performance score of the currently active link and the preset switching hysteresis threshold.

2. The method for managing multiple communication links in a wireless ad hoc network according to claim 1, characterized in that, The step of generating the predictive occlusion index includes: S21. Based on the situational data, calculate the line-of-sight occlusion component, which represents the possibility that the line-of-sight path is physically blocked. S22. Based on the situational data, calculate the diffraction influence component that characterizes the proximity of the line-of-sight path to the dynamic obstacle; S23. Based on the preset risk model weights, the line-of-sight occlusion component and the diffraction influence component are weighted and combined to generate the predictive occlusion index.

3. The method for managing multiple communication links in a wireless ad hoc network according to claim 1, characterized in that, The steps for calculating the overall link performance score include: S31. Convert the predictive occlusion index generated in step S2 into a risk aversion score; S32. Calculate the normalized service quality score based on the real-time service quality parameters; S33. Based on the preset operating cost data, calculate the normalized inverse cost score; S34. Based on the preset strategy weights, the risk avoidance score, the normalized service quality score, and the normalized inverse cost score are weighted and summed to generate the link comprehensive performance score.

4. The method for managing multiple communication links in a wireless ad hoc network according to claim 3, characterized in that, The steps for calculating the normalized service quality score include: S321. Calculate the original service quality score by combining the link latency and jitter in the real-time service quality parameters according to the preset weight combination. S322. Using the historical maximum and minimum raw service quality scores observed within a preset time period, normalize the current raw service quality score to generate the normalized service quality score.

5. The method for managing multiple communication links in a wireless ad hoc network according to claim 3, characterized in that, The steps for calculating the normalized inverse cost score include: S331. Retrieve the cost value corresponding to the current link from the preset cost table; S332. The retrieved cost value is normalized using the highest cost value among all available links to generate the normalized inverse cost score.

6. The method for managing multiple communication links in a wireless ad hoc network according to claim 3, characterized in that, The preset strategy weights can be configured and adjusted between cost-sensitive mode, security-first mode, and performance-first mode according to operational business objectives.

7. The method for managing multiple communication links in a wireless ad hoc network according to claim 1, characterized in that, The duration of the future prediction time window is set to be greater than the total time required for the system to execute the complete link switching process, so as to ensure that a safety margin is reserved for the switching execution before the predicted link interruption actually occurs.

8. A method for managing multiple communication links in a wireless ad hoc network according to claim 2, characterized in that, The diffraction effect component is calculated based on the shortest distance between the line-of-sight path and all dynamic obstacles, and is determined with reference to a preset safe distance threshold; the safe distance threshold is determined based on the Fresnel zone radius corresponding to the operating frequency of the communication link.

9. A multi-communication link management platform for a wireless ad hoc network, employing the multi-communication link management method for a wireless ad hoc network as described in any one of claims 1-8, characterized in that, include: The situational awareness module is used to collect the real-time three-dimensional position coordinates, velocity vectors and predetermined driving trajectories of unmanned vehicle nodes in the network, and to obtain the real-time position and velocity of other unmanned vehicle nodes in the network as dynamic obstacle data. Combined with a preset three-dimensional digital map, situational data is generated. The link occlusion risk prediction module is used to quantitatively predict the occlusion risk of the available communication links of the unmanned vehicle node within a preset future prediction time window based on the situation data generated by the situation awareness module, and generate a predictive occlusion index. The multi-link comprehensive performance evaluation module is used to receive the predictive obstruction index generated by the link obstruction risk prediction module, and combine it with the real-time service quality parameters obtained from the network protocol stack and the preset operating cost data to calculate and output the comprehensive performance score of the available communication links. A predictive switching decision and execution module is used to switch the service data flow from the currently active link to the inactive link with the highest overall link performance score in response to the link performance score meeting a preset switching condition. The preset switching condition is that the inactive link with the highest overall link performance score, after a preset stable period of time, is still higher than the sum of the overall link performance score of the currently active link and a preset switching hysteresis threshold.

Images