Multi-hop communication node intelligent scheduling method and system

Abstract

The invention discloses a multi-hop communication node intelligent scheduling method and system, and relates to the technical field of communication networks. The method is used for solving the scheduling and switching problems of multi-hop relay nodes in disaster emergency communication. Residual energy, channel fading parameters and link load data of an unmanned aerial vehicle and a ground terminal are collected, a communication potential index is calculated in combination with dynamic weight fusion, and the relay capability of a node in a complex environment is accurately evaluated. A relay node selection threshold is set according to the disaster type, nodes with communication potential indexes higher than the disaster adaptation threshold are preferentially selected, an initial hybrid topology structure with preferential survivability is constructed, and the network stability and survivability are improved. An autonomous game agent model is adopted, a multi-target reward function is designed based on local channel quality, residual energy and task emergency degree, it is ensured that the network achieves collaborative optimization of energy consumption balance, transmission real-time performance and survivability under the condition of dynamic topology and node failure, and efficient and continuous optimization of the disaster emergency communication network is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of communication networks, and particularly to an intelligent scheduling method and system for multi-hop communication nodes. Background Art

[0002] In sudden disasters such as earthquakes, floods, and typhoons, traditional emergency communication networks are often restricted in emergency response and rescue operations due to damaged or unrecoverable infrastructure in a timely manner. Especially after a disaster occurs, in remote areas or disaster-stricken areas with complex environments, communication links are easily affected by multiple factors such as channel fading, network congestion, and energy consumption. To ensure smooth communication during a disaster and provide timely and effective rescue information, an intelligent multi-hop communication network has become one of the important emergency communication means. Especially with the popularization and development of unmanned aerial vehicles and ground emergency terminals, combined with wireless communication technology and intelligent scheduling methods, a self-organizing network can be quickly established after a disaster to provide flexible and efficient communication support.

[0003] Currently, the application of traditional multi-hop communication methods in emergency communication still faces multiple challenges. First, existing methods often ignore the changes in the remaining energy and channel quality of nodes in a disaster environment, resulting in the network being easily interrupted due to energy exhaustion or unstable channels in sudden situations. Second, traditional scheduling strategies mostly rely on static routing or simple load balancing algorithms, and do not fully consider the differences in disaster types and the real-time status of nodes. Third, existing methods lack an adaptive mechanism to cope with network topology changes and node failures. Especially in a dynamic and complex post-disaster environment, they cannot adjust relay nodes or communication links in a timely manner, affecting the stability and real-time performance of communication. Finally, traditional methods are mostly limited to the optimization of a single objective, such as energy or delay, and ignore the collaborative optimization of multiple objectives, making it difficult to achieve a balance among energy efficiency, transmission timeliness, and network survivability in a complex environment. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an intelligent scheduling method and system for multi-hop communication nodes, which solves the problems in the above background art.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: A multi-hop communication node intelligent scheduling method, comprising the following steps: S1. Real-time collect the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal communication nodes, and calculate the communication potential index through dynamic weight fusion, which is used to comprehensively quantify the multi-hop relay ability of the nodes under energy constraints and channel mutations; S2. According to the communication potential index, set the relay node selection threshold for different disaster types, and preferentially select the communication nodes with a communication potential index higher than the disaster adaptation threshold as relay nodes to construct an initial hybrid topology structure with priority for anti-destruction, and combine the disaster heat distribution map to perform load pre-allocation on the key rescue data streams; S3. When it is detected that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to sudden energy drop or channel perturbation, trigger the role migration operation, broadcast a migration request carrying an emergency level identifier to neighboring nodes, and select the optimal candidate node to take over the relay function according to the real-time communication potential index sorting result and link survival prediction information; S4. Model each node as an autonomous game agent, whose action space includes four states: relay, terminal, standby, and emergency communication. Design a multi-objective reward function according to the local channel quality, remaining energy, and urgency of the rescue task, and incrementally game decision-making information through anti-interference beacon interaction, and converge to the Nash equilibrium state under dynamic topology and partial node failure scenarios, generating a multi-hop scheduling strategy that collaboratively optimizes energy consumption balance, transmission real-time performance, and network anti-destruction.

[0006] Further, the specific process of real-time collecting the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal communication nodes is as follows: Through the energy monitoring modules deployed on the UAV and ground terminals, periodically collect the remaining battery energy value, and combine the charge and discharge rate to predict the energy decay trend; Through the real-time communication feedback mechanism at the link layer, count the received signal strength and bit error rate between each pair of communication nodes, calculate their instantaneous channel fading level, and perform sliding analysis on the fluctuation amplitude within a short time to extract the channel quality change gradient to determine whether there is a risk of channel mutation; Combine the forwarding queue status and actual load situation of each node, and real-time monitor the buffer queue length, average waiting delay, and packet queuing frequency to form a load vector reflecting the link load level, and normalize the load peak of different time slices to generate a standardized link pressure eigenvalue.

[0007] Furthermore, the specific process of calculating the communication potential index through dynamic weight fusion is as follows: For each communication node, standardized index values are extracted from three dimensions: remaining energy, channel fading, and link load, and their numerical ranges are unified to avoid calculation biases caused by inconsistent dimensions between different dimensions. According to the current disaster type, node role tasks, and task urgency parameters, different dynamic weights are assigned to the three indicators to generate a weight coefficient group reflecting the current scheduling priority requirements. The three weighted indicators are fused to construct a three-dimensional feature vector, and by calculating the mapping function value in a specific scheduling strategy space, a communication potential index representing the multi-hop relay ability of the current node is generated. Deviation analysis is performed on the distribution of the communication potential index among all candidate nodes to determine whether the node has a significant relay ability advantage, which serves as an important input signal for subsequent relay selection and migration criteria.

[0008] Furthermore, according to the communication potential index, the specific process of setting the relay node selection threshold for different disaster types is as follows: Obtain the communication guarantee level requirements corresponding to the current disaster type, and based on the distribution range of the communication potential index in historical disaster emergency communication tasks, extract the minimum communication potential index value that successfully maintains the stability of the communication link in various disaster scenarios as the reference threshold. Combine the current task urgency and node distribution density to set a weighted correction coefficient to dynamically adjust the basic threshold to form a disaster situation-adaptive communication potential selection threshold. Sort the communication potential indexes of all candidate nodes and select the nodes with indexes higher than the set threshold as the alternative set of relay nodes. On the basis of the alternative node set, further screen the target nodes suitable as relays in combination with the following specific criteria: Geographic location constraint: Exclude nodes located in communication blind spots, with severe terrain occlusion, or more than the preset communication radius from the target service area; Link connectivity constraint: Select nodes with the number of adjacent nodes higher than the average connection degree threshold through the adjacency matrix calculation method; Energy distribution balance constraint: Calculate the energy variance of the alternative node set, preferentially select nodes with battery levels in the medium to high level range, and exclude nodes with the current remaining energy lower than a certain proportion of the network average value.

[0009] Furthermore, the specific process of constructing an initial hybrid topology with priority on anti-destruction and pre-assigning loads to critical rescue data streams in combination with the disaster heat distribution map is as follows: Select backbone nodes through the communication potential index, where some backbone nodes serve as cluster head nodes to cover high-priority rescue areas, and the remaining backbone nodes construct a mesh redundant layer to form a communication topology with multi-path connection capabilities; Manage rescue data hierarchically, allocate different transmission resources and path redundancy levels according to the importance of the data, where high-priority data is configured with fixed time slots and multi-path transmission strategies, low-priority data uses dynamic resource scheduling methods, and backup channels are pre-configured for critical data to enhance link reliability; Deploy redundant cluster head nodes in potentially high-risk areas to form an overlapping coverage structure, and set up a load awareness mechanism at critical nodes to dynamically migrate some traffic to neighboring nodes according to the current load status of the nodes to achieve link load balancing; The topology drives each node to update its local routing information in real time by periodically broadcasting topology status beacons, and when a link is interrupted, it quickly selects alternative nodes based on the communication potential index to complete path reconstruction.

[0010] Furthermore, the specific process of selecting the optimal candidate node to take over the relay function based on the real-time communication potential index sorting results and link survival prediction information is as follows: Receive the real-time communication potential index broadcast by neighboring nodes, generate a candidate node list in descending order, and combine geographical location and communication range limitations to eliminate nodes beyond the effective connection range; Based on the historical channel stability data of candidate nodes and the link survival prediction model, screen nodes with a survival probability higher than the set threshold to form a highly reliable candidate set; Send a migration request to the nodes in the highly reliable candidate set and verify their current load status to ensure that their communication potential index and load level meet the takeover conditions; Achieve seamless switching through the dual-buffer routing table mechanism, where the old table maintains the existing data transmission, and the new table completes the migration and takes effect within the set time; Trigger the broadcast of local topology synchronization beacons to update the routing information of all network nodes, and quickly reconstruct the path based on the communication potential index when a link is interrupted.

[0011] Furthermore, a multi-objective reward function is designed based on local channel quality, remaining energy, and the urgency of rescue tasks. The specific process of converging to the Nash equilibrium state through anti-jamming beacon interaction incremental game decision-making information in the scenarios of dynamic topology and partial node failures is as follows: According to local channel quality, remaining energy, and the urgency of rescue tasks, a channel gain term, an energy penalty term, and a task priority weighting term are designed respectively, and a multi-objective composite reward function is constructed; Local decision-making information, including node action strategies, reward value increments, and link state summaries, is compressed and encoded by anti-jamming beacons and broadcast to neighboring nodes using frequency-hopping spread-spectrum technology; Nodes update their local strategies based on the received game information, and iteratively optimize the action selection probability distribution through distributed reinforcement learning; When the strategy difference degree of all network nodes continues to be lower than the set threshold and the key performance indicators tend to be stable, it is determined that the system converges to the Nash equilibrium state.

[0012] Furthermore, the specific process of generating a multi-hop scheduling strategy for collaborative optimization of energy consumption balance, transmission real-time performance, and network survivability is as follows: Relay tasks are dynamically allocated based on the Nash equilibrium strategy, the continuous working duration of high-load nodes is restricted, and low-load nodes are triggered to take over part of the forwarding traffic; Dedicated low time slots are reserved for key data streams of vital signs and location coordinates, allowing them to preempt non-critical data time slots to ensure transmission real-time performance; Redundant paths are deployed in potential risk areas of the disaster heat map, and the optimal backup link is quickly activated based on the difference in communication potential index when the main path is interrupted; The whole network topology state synchronization is periodically triggered, the node routing table is dynamically updated, and local congestion diffusion is suppressed to ensure network survivability.

[0013] A multi-hop communication node intelligent scheduling system includes the following modules: a communication potential analysis module, a relay node selection module, a role migration trigger module, and an autonomous game decision-making module. The communication potential analysis module is used to collect the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal communication nodes in real time, and calculate the communication potential index through dynamic weight fusion, which is used to comprehensively quantify the multi-hop relay ability of the nodes under energy limitation and channel mutation. The relay node selection module is used to set the relay node selection threshold for different disaster types according to the communication potential index, preferentially select the communication nodes with a communication potential index higher than the disaster-adapted threshold as relay nodes, construct an initial hybrid topology with priority given to anti-destruction, and perform load pre-allocation on the key rescue data streams in combination with the disaster heat distribution map. The role migration trigger module is used to trigger the role migration operation when it detects that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to sudden energy drop or channel perturbation, broadcast a migration request carrying an emergency level identifier to neighboring nodes, and select the optimal candidate node to take over the relay function according to the real-time communication potential index sorting result and link survival prediction information. The autonomous game decision-making module is used to model each node as an autonomous game agent, whose action space includes four states: relay, terminal, standby, and emergency communication. A multi-objective reward function is designed according to the local channel quality, remaining energy, and urgency of the rescue task, and incremental game decision-making information is interacted through anti-interference beacons, converging to the Nash equilibrium state in the dynamic topology and partial node failure scenarios, and generating a multi-hop scheduling strategy that collaboratively optimizes energy consumption balance, transmission real-time performance, and network anti-destruction ability.

[0014] The present invention has the following beneficial effects: (1) A multi-hop communication node intelligent scheduling method. By collecting the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal in real time and combining the communication potential index calculated by dynamic weight fusion, the present invention can comprehensively quantify the multi-hop relay ability of the nodes under energy limitation and channel mutation. This enables the network to adaptively select the most suitable relay nodes in the emergency communication scenario, ensuring the stability and reliability of data transmission. At the same time, setting the relay node selection threshold for different disaster types according to the communication potential index effectively constructs a hybrid topology with priority given to anti-destruction, and reasonably pre-allocates the load of the key rescue data streams, improving the network's disaster response ability.

[0015] (2) A multi-hop communication node intelligent scheduling system. By introducing an intelligent agent model based on autonomous game and combining a multi-objective reward function designed according to local channel quality, remaining energy, and the urgency of rescue tasks, in an environment of dynamic topology and partial node failures, the present invention can ensure that the system quickly responds and converges to the Nash equilibrium state, optimizing the energy balance, transmission real-time performance, and network survivability of multi-hop communication. In addition, when a node fails, the system can quickly reconstruct the path and synchronize the network topology, ensuring the continuity and stability of communication and enhancing the disaster recovery ability of the entire network.

[0016] Of course, it is not necessary for any product implementing the present invention to achieve all the above-mentioned advantages simultaneously. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a flowchart of an intelligent scheduling method for multi-hop communication nodes of the present invention.

[0018] Figure 2 It is a flowchart of an intelligent scheduling system for multi-hop communication nodes of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In the embodiments of the present application, through an intelligent scheduling method and system for multi-hop communication nodes, by collecting the energy, channel state, and link load data of nodes in real time and calculating the communication potential index through dynamic weights, the scheduling and switching problems of multi-hop relay nodes in disaster emergency communication are solved. This method can preferentially select appropriate relay nodes according to the type of disaster and trigger role migration when the energy or channel state of the node changes, ensuring the stability and efficiency of the network in a complex environment.

[0020] The general idea of the solution in the embodiments of the present application is as follows: Collect the remaining energy, channel fading parameters, and link load data of the communication nodes of the unmanned aerial vehicle and the ground emergency terminal in real time, and calculate the communication potential index through dynamic weight fusion, which is used to comprehensively quantify the multi-hop relay ability of the node under energy constraint and channel mutation.

[0021] According to the communication potential index, set the relay node selection threshold for different types of disasters, preferentially select the communication nodes with a communication potential index higher than the disaster-adapted threshold as relay nodes, construct an initial hybrid topology structure with priority on survivability, and perform load pre-allocation on the key rescue data streams in combination with the disaster heat distribution map.

[0022] When it is detected that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to a sudden drop in energy or channel perturbation, trigger the role migration operation, broadcast a migration request carrying an emergency level identifier to neighboring nodes, and select the optimal candidate node to take over the relay function according to the sorting result of the real-time communication potential index and the link survival prediction information.

[0023] Model each node as an autonomous game agent. Its action space includes four states: relay, terminal, standby, and emergency communication. Design a multi-objective reward function based on local channel quality, remaining energy, and the urgency of rescue tasks. Incrementally game decision-making information through anti-interference beacon interaction, and converge to the Nash equilibrium state in the scenarios of dynamic topology and partial node failures, generating a multi-hop scheduling strategy that collaboratively optimizes energy consumption balance, transmission real-time performance, and network survivability.

[0024] Please refer to Figure 1 , an embodiment of the present invention provides a technical solution: a multi-hop communication node intelligent scheduling method, including the following steps: S1. Real-time collect the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal communication nodes, and calculate the communication potential index through dynamic weight fusion, which is used to comprehensively quantify the multi-hop relay ability of the node under energy limitation and channel mutation; S2. According to the communication potential index, set the relay node selection threshold for different disaster types, and preferentially select the communication nodes with a communication potential index higher than the disaster adaptation threshold as relay nodes to construct an initial hybrid topology structure with priority on survivability, and combine the disaster heat map to pre-allocate the load of key rescue data streams; S3. When it is detected that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to sudden energy drop or channel perturbation, trigger the role migration operation, broadcast a migration request carrying an emergency level identifier to neighboring nodes, and select the optimal candidate node to take over the relay function according to the real-time communication potential index sorting result and link survival prediction information; S4. Model each node as an autonomous game agent. Its action space includes four states: relay, terminal, standby, and emergency communication. Design a multi-objective reward function based on local channel quality, remaining energy, and the urgency of rescue tasks. Incrementally game decision-making information through anti-interference beacon interaction, and converge to the Nash equilibrium state in the scenarios of dynamic topology and partial node failures, generating a multi-hop scheduling strategy that collaboratively optimizes energy consumption balance, transmission real-time performance, and network survivability.

[0025] In this implementation scheme, step S1: The system real-time obtains the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal. The collection of these data is used to calculate the communication potential index.

[0026] Residual energy: Refers to the remaining battery energy of the drone or ground terminal at present. The higher the energy, the stronger the communication ability of the node. Channel fading parameter: Describes the attenuation of wireless communication signals during transmission due to environmental factors (such as weather, obstacles, etc.). Poor channel quality will affect the stability of communication. Link load data: Represents the load situation of the current node's communication, that is, whether the node is in an overloaded state. Nodes with excessive load may not be able to continue to forward data efficiently. Through these data, the system uses dynamic weight fusion calculation to obtain the communication potential index of each node. This index combines the residual energy, channel quality, and load situation of the node, and is used to measure the multi-hop relay ability of the node under energy-constrained and channel-fluctuating conditions. Step S2: According to the communication potential index calculated in the previous step, the system sets a relay node selection threshold for different disaster types. Disaster type: Different disasters (such as earthquakes, floods, etc.) have different requirements for the communication network. The system sets the threshold according to the characteristics of the disaster to ensure the selection of the most suitable nodes for relaying. Nodes with a communication potential index higher than the disaster-adapted threshold: Under disaster conditions, the system preferentially selects those nodes with a communication potential index higher than the set threshold as relay nodes because these nodes can provide more stable communication services. Then, the system constructs an initial hybrid topology with priority given to survivability, that is, according to the communication potential index of the nodes, nodes with stronger survivability are selected to ensure that the communication network can maintain better stability in disasters. In addition, the system also performs load pre-allocation for critical rescue data streams based on the disaster heat map, that is, reasonably distributes data traffic according to the severity and distribution of the disaster to avoid communication bottlenecks. Step S3: When it is monitored that the communication potential index of a certain relay node is lower than the preset safety threshold due to a sudden drop in energy or channel perturbation, the system will initiate a role migration operation. Role migration operation: Refers to migrating the task of the node that originally undertook the relay task to other communication nodes to maintain the stability and reliability of the communication network. Emergency level identifier: Each migration request will carry an emergency level identifier, indicating the urgency of the current request. Requests with a higher priority will be processed first. The system broadcasts a migration request, requesting neighboring nodes to take over the relay function, and selects the optimal candidate node according to the real-time communication potential index and link survival prediction information. This step ensures that when the original node fails, the network can quickly switch to the best relay node to avoid communication interruption. Step S4: In this step, the system models each communication node as an autonomous game agent. This means that each node makes decisions independently, and the nodes interact through games to jointly optimize the operation of the network. Autonomous game agent: When making decisions, each node decides whether to act as a relay node, a terminal, a standby node, or an emergency communication node based on local information and environmental factors (such as channel quality, residual energy, task urgency).Action space: It refers to the types of operations that each node can choose, such as "relay" (forwarding data), "terminal" (receiving or sending data), "standby" (in a standby state), or "emergency communication" (processing emergency tasks). Based on these factors, each node designs a multi-objective reward function that comprehensively considers the local channel quality, remaining energy, and task urgency of the node, ensuring that under different conditions, the network can optimize energy consumption, real-time transmission ability, and survivability. Nash equilibrium: Through game decision-making information, the system expects to achieve a Nash equilibrium among nodes, that is, the decision of each node will not change given the decisions of other nodes. In this way, each node in the network will automatically adjust to an optimized state, maintaining the best operating state of the system in the case of network topology changes and partial node failures. Ultimately, these decisions will lead to the collaborative optimization of energy consumption balance, transmission real-time performance, and survivability, thus generating an efficient and reliable multi-hop scheduling strategy to ensure the stability and real-time performance of communication in a disaster emergency communication environment.

[0027] Specifically, the specific process of real-time collecting the remaining energy, channel fading parameters, and link load data of UAVs and ground emergency terminal communication nodes is as follows: Through the energy monitoring modules deployed on UAVs and ground terminals, the remaining battery energy values are collected periodically, and the energy decay trend is predicted by combining the charge and discharge rates; Through the real-time communication feedback mechanism at the link layer, the received signal strength and bit error rate between each pair of communication nodes are statistically analyzed, the instantaneous channel fading level is calculated, and the sliding analysis of the fluctuation amplitude within a short time is performed to extract the channel quality change gradient to determine whether there is a risk of channel mutation; Combining the forwarding queue status and actual load conditions of each node, the buffer queue length, average waiting delay, and packet queuing frequency are monitored in real time to form a load vector reflecting the link load level, and the load peak values in different time slices are normalized to generate a standardized link pressure eigenvalue.

[0028] In this implementation scheme, the process of deploying the energy monitoring module to collect the remaining energy value of the battery: The energy monitoring module is a hardware device installed on the UAV and the ground emergency terminal, responsible for periodically collecting the remaining energy value of the battery. These devices monitor the battery status through the built-in battery management system. The remaining energy value of the battery represents the power that the current battery can provide. The remaining battery energy is crucial for determining whether the node can continue to undertake communication tasks. Charge and discharge rate: During the operation of the battery, there are charging and discharging processes. The monitoring module combines the charge and discharge rate of the battery to predict the attenuation trend of the battery in the next period of time. For example, if the battery is in a high discharge rate state, it may cause the power to be consumed quickly, affecting the working time of the node. Prediction of energy attenuation trend: By predicting the energy consumption speed of the battery, decisions can be made in advance to avoid the node being unable to continue communication relay due to battery depletion. The channel fading parameters are statistically analyzed through the real-time communication feedback mechanism at the link layer. Process: Real-time communication feedback mechanism at the link layer: This is a feedback mechanism in wireless communication, which means that during data transmission, communication nodes feedback information such as signal strength and bit error rate through the link layer. Through the feedback information, the system can judge the change of channel quality. Received signal strength: It represents the signal strength received by the receiving end of the communication link from the sending end. A higher signal strength indicates good channel quality and high communication stability, otherwise it may cause data loss or errors. Bit error rate: It represents the ratio of errors occurring during data transmission. The lower the bit error rate, the better the channel quality and the higher the reliability of data transmission. Instantaneous channel fading level: Channel fading refers to the attenuation of the signal during propagation due to factors such as multipath propagation and interference. The instantaneous channel fading level refers to the fading situation of the channel at a certain moment. The system can evaluate the current channel quality by measuring this level. Process of sliding analysis of the change gradient of channel quality: Sliding analysis: To accurately capture the change trend of channel fading, the sliding window method is used to analyze the changes in signal strength or bit error rate. This method can smooth the signal fluctuations, reduce the impact of instantaneous changes, and more accurately reflect the long-term trend. Change gradient of channel quality: It represents the speed or amplitude of the change of channel quality over time. For example, if the channel quality drops sharply in a very short time, it may mean that the channel is undergoing a mutation. By calculating the change gradient of channel quality, the system can give an early warning of potential channel mutation problems. Process of monitoring the link load by combining the status of the node forwarding queue: Status of the forwarding queue: When each communication node processes data, it may need to forward the data to other nodes. The forwarding queue of the node is used to cache the data packets to be forwarded. The queue status reflects the current load situation of the node. Length of the cache queue: It refers to the number of data packets stored in the current cache queue. If the queue length is long, it means that the node has a heavy load, and queuing waiting problems may occur, resulting in data forwarding delay. Average waiting delay: The average time for a data packet to enter the queue and be forwarded out.Long waiting times may affect the real-time nature of data, especially in emergency communications where timely data forwarding is crucial. Packet queuing frequency: This represents the frequency at which packets enter the queue. If the queuing frequency is too high, it may lead to an increase in communication latency at the node. Process of generating a standardized link stress eigenvalue: Load vector: Combining the cache queue length, average waiting delay, and packet queuing frequency, the system generates a load vector that contains multiple parameters representing the node's load. These parameters are used to evaluate whether the node is overloaded. Normalization of the load peak: The load level of the node fluctuates over time, so it is necessary to normalize the load peak for each time slice. The normalization process transforms the load data of different time slices to a unified scale, avoiding biases caused by time differences and making the load characteristics more comparable. Link stress eigenvalue: The normalized load vector can generate a standardized link stress eigenvalue, which is used to quantify the load pressure on the current link. This eigenvalue can help the system determine the congestion level of the link, thereby deciding whether to adjust the data forwarding strategy or whether to migrate the relay task to a node with a lighter load.

[0029] Specifically, the specific process of calculating the communication potential index through dynamic weight fusion is as follows: For each communication node, standardized index values are extracted from three dimensions: remaining energy, channel fading, and link load, and their numerical ranges are unified to avoid calculation biases caused by inconsistent dimensions between them; according to the current disaster type, node role tasks, and task urgency parameters, different dynamic weights are assigned to the three indicators to generate a weight coefficient group reflecting the current scheduling priority requirements; the three weighted indicators are fused to construct a three-dimensional feature vector, and by calculating the mapping function value in a specific scheduling strategy space, a communication potential index representing the multi-hop relay ability of the current node is generated; deviation analysis is performed on the distribution of the communication potential index among all candidate nodes to determine whether the node has a significant relay ability advantage, which is used as an important input signal for subsequent relay selection and migration criteria.

[0030] In this implementation plan, the purpose of multi-dimensional data normalization processing: Eliminate the dimensional differences of the three dimensions of remaining energy, channel fading, and link load to ensure data comparability. Normalization processing is performed on the data of each dimension separately, mapping it to the interval [0, 1]. Formula representation: Parameter description: : The original data of the kth dimension ( is the remaining energy, is the channel fading, is the link load); 、 : The historical minimum and maximum values of the k-th dimension. Dynamic weight allocation: Dynamically adjust the weights of each dimension according to the disaster type, node role, and task urgency. Define the weight allocation rules: Earthquake scenario: Increase the energy weight (emphasize node survival ability); Flood scenario: Enhance the channel fading weight (suppress the impact of multipath interference); Fire scenario: Increase the link load weight (avoid overloading nodes in high-temperature areas). Formula representation: ; Parameter description: : The weight coefficients of remaining energy, channel fading, and link load (satisfying ); : Disaster type code (e.g., earthquake = 1, flood = 2, fire = 3); : Weight allocation function, dynamically output the weight value according to the disaster type. Three-dimensional feature vector fusion: Fuse the weighted multi-dimensional indicators into a single communication potential index (CPI). Perform weighted summation on the standardized indicators and enhance the discrimination through non-linear mapping. Formula representation: ; Parameter description: : Sigmoid function ( ), used to compress the result into the interval [0, 1]; : Link load reverse processing (the lower the load, the higher the contribution value). Deviation analysis and relay advantage determination Purpose: Screen nodes with significant relay ability advantages. Calculate the mean value ( ) and standard deviation ( ) of the CPI of all candidate nodes; Judge the relative advantage of the node through Zscore standardization: ; If , determine that the node has a significant relay advantage and give priority to selecting it as a candidate. Disaster scenario adaptability: Through the dynamic weight allocation function ( ), flexibly adapt to the core requirements of scenarios such as earthquakes, floods, and fires; Anti-deviation design: Normalization processing ( ) and non-linear mapping (Sigmoid) suppress the interference of extreme values; Interpretability: Zscore quantifies the relay ability advantage of the node and provides a clear criterion for migration decision-making.

[0031] Specifically, according to the communication potential index, the specific process of setting the selection threshold of relay nodes for different disaster types is as follows: Obtain the communication guarantee level requirements corresponding to the current disaster type, and extract the minimum communication potential index that successfully maintains the stability of the communication link in various disaster scenarios as the reference threshold according to the distribution range of the communication potential index in historical disaster emergency communication tasks; Combine the current task urgency and node distribution density to set a weighted correction coefficient to dynamically adjust the basic threshold to form a communication potential selection threshold adapted to the disaster situation; Sort the communication potential indices of all candidate nodes, and screen out the nodes with communication potential indices higher than the set threshold as the alternative set of relay nodes; On the basis of the alternative node set, further screen the target nodes suitable as relays in combination with the following specific criteria: Geographic location constraint: Eliminate the nodes located in communication blind spots, with severe terrain occlusion, or exceeding the preset communication radius from the target service area; Link connectivity constraint: Select the nodes with the number of adjacent nodes higher than the average connection degree threshold through the adjacency matrix calculation method; Energy distribution balance constraint: Calculate the energy variance of the alternative node set, give priority to selecting the nodes with the battery level in the medium-high level range, and eliminate the nodes with the current remaining energy lower than a certain proportion of the network average value.

[0032] In this implementation plan, the type of disaster usually affects the requirements of the communication network. For example, in natural disasters, different service qualities are required for the guarantee level of the communication network. For instance, the more severe the disaster, the higher the required network reliability and stability. Based on different disaster types, the system will obtain the communication guarantee level requirements in the disaster scenario. These requirements will determine the selection criteria for relay nodes in the network. For example, some disaster types may have very high requirements for the stability of communication links, while others have lower requirements. The distribution data of the communication potential index of historical disaster emergency communication tasks can be used as a reference to determine the minimum communication potential index for successfully maintaining the stability of communication links in similar disaster scenarios. The system will count the minimum communication potential index of the nodes that have successfully maintained communication based on the data of historical disaster events, and this value is used as a reference threshold to set the selection criteria in the current task. After obtaining the basic reference threshold, the system will perform weighted correction according to the current task urgency and node distribution density. Task urgency: Tasks with high urgency may require stronger potential of communication nodes, so the basic threshold will be increased to ensure that the selected nodes have sufficient capabilities to support the task. Node distribution density: The distribution density of nodes may affect the node selection strategy. In areas with dense nodes, the threshold can be appropriately reduced, while in areas with sparse nodes, a higher potential value is required to ensure the stability of the network. Based on these factors, the basic threshold is dynamically adjusted to ensure that the threshold can adaptively adapt to the current disaster situation and task requirements, forming a disaster situation adaptive communication potential selection threshold. According to the aforementioned adaptive threshold, the communication potential indexes of all candidate nodes are sorted. The sorting process is to arrange the communication potential indexes of all nodes in descending order, and select the nodes with communication potential indexes higher than the set threshold as the alternative set of relay nodes. At this time, the alternative nodes already meet at least the basic communication potential requirements and become potential relay node candidates. After initially screening out the nodes that meet the threshold conditions, several other specific criteria need to be combined to further screen out the target nodes suitable as relays. The specific criteria include: Geographic location constraint: Exclude the nodes located in communication blind spots or with severe terrain occlusion. For example, some nodes may be blocked by mountains or buildings, and the signal propagation is greatly affected. These nodes are not suitable as relay nodes. Nodes located more than the preset communication radius from the target service area also need to be excluded. That is, when the node is too far from the disaster area, it may not be able to provide stable communication services, so it is also excluded. Link connectivity constraint: Use the adjacency matrix to calculate the connectivity of each candidate node. The adjacency matrix represents the connection relationship between nodes, and select the nodes with strong connection relationships with other nodes, which can ensure the connectivity of relay nodes in the network. According to the number of adjacent nodes of the node, select the nodes with the number of adjacent nodes higher than the average connection degree threshold, that is, select the nodes directly connected to more nodes to improve the robustness of the network. Energy distribution balance constraint: Calculate the energy variance of the alternative node set, that is, the degree of dispersion of node energy.Through energy variance analysis, nodes with overly dispersed energy can be avoided, as these nodes may not be able to support relay tasks for a long time. Nodes with medium to high levels of power are preferentially selected to ensure that relay nodes do not stop working due to energy exhaustion during task execution. Nodes with remaining energy below a certain percentage of the network-wide average are excluded to ensure the stability of relay nodes, avoid over-reliance on nodes with low power, and guarantee the continuity of the communication network.

[0033] Specifically, the specific process of constructing an initial hybrid topology with priority on survivability and pre-assigning loads to critical rescue data streams in combination with the disaster heat distribution map is as follows: backbone nodes are selected through the communication potential index, with some of the backbone nodes serving as cluster head nodes to cover high-priority rescue areas, and the remaining backbone nodes constructing a mesh redundancy layer to form a communication topology with multi-path connection capabilities; the rescue data is hierarchically managed, and different transmission resources and path redundancy levels are allocated according to the importance of the data. Among them, high-priority data is configured with fixed time slots and multi-path transmission strategies, while low-priority data uses dynamic resource scheduling methods, and backup channels are pre-configured for critical data to enhance link reliability; redundant cluster head nodes are deployed in potentially high-risk areas to form an overlapping coverage structure, and a load awareness mechanism is set at critical nodes to dynamically migrate some traffic to neighboring nodes according to the current load status of the nodes to achieve link load balancing; the topology broadcasts topology status beacons periodically to drive each node to update its local routing information in real time, and when a link is interrupted, alternative nodes are quickly selected according to the communication potential index to complete path reconstruction.

[0034] In this implementation plan: Backbone nodes: These are the core nodes in the network, undertaking key relay tasks. The selection of backbone nodes is based on the communication potential index, which is the comprehensive ability of nodes in terms of energy, channel quality, load, etc. Selecting nodes with a higher communication potential index can ensure that these nodes have strong relay and routing capabilities. Cluster head nodes: Some backbone nodes are selected as cluster head nodes. Cluster head nodes are responsible for managing the communication within their coverage area, mainly for high-priority rescue areas. Mesh redundancy layer: The remaining backbone nodes form a mesh redundancy layer, that is, these nodes are connected to other nodes through multi-path methods to form redundant connections, ensuring that the network can still maintain stable communication when some nodes or links fail. The mesh topology has stronger survivability because even if some links or nodes fail, data can be transmitted through other paths. Data hierarchical management: Different rescue data are assigned different priorities according to their importance. For high-priority data (such as emergency rescue instructions, vital sign data, etc.), the system will allocate higher transmission resources and path redundancy levels to ensure that this data can be transmitted timely and reliably. Low-priority data (such as regular status reports, etc.) adopts a dynamic resource scheduling method, flexibly adjusting the transmission bandwidth and path according to the current network load situation. Multi-path transmission strategy: For high-priority data, in addition to fixed transmission time slots, a multi-path transmission strategy is also adopted, that is, the same data stream is transmitted through different paths, thereby enhancing the reliability and fault tolerance of data transmission. Backup channels: Provide backup channels for critical data streams to enhance the reliability of the links. Even if a certain channel is interrupted, the system can continue to transmit data through the backup channel, avoiding the negative impact of communication interruption on the rescue mission. Redundant cluster head nodes: To improve the survivability in high-risk areas, the system will deploy multiple redundant cluster head nodes in potential high-risk areas (such as the core area of the disaster site). These nodes will cover the area and ensure that the communication links within the area will not be interrupted due to node failure through cross-coverage. Cross-coverage structure: The cross-coverage structure between redundant cluster head nodes is to avoid affecting the communication of the entire area when a single cluster head node fails. Through redundancy and cross-coverage, the network can quickly restore communication when node or link failures occur. Load awareness mechanism: Each node judges whether the current node is overloaded by monitoring its own load status (such as queue length, transmission rate, etc.). If a node has a high load, the system will dynamically migrate some traffic to neighboring nodes with lower loads. This mechanism helps with load balancing, ensuring that the data streams in the network can be reasonably distributed among nodes and preventing a single node from being overloaded, which may lead to network congestion or communication interruption. Link load balancing: Through the load awareness mechanism and traffic migration, the system can adjust the distribution of data streams in real time under dynamically changing network conditions, ensuring the stability and efficiency of communication links.Periodic Broadcast of Topology State Beacons: Each node periodically broadcasts its own topology state information. These signals contain information such as the communication status and connection status of the node. Other nodes update their local routing information in real time by receiving these signals. In this way, the nodes in the network can understand the changes in the global topology and make routing decisions accordingly. Path Reconstruction during Link Failure: In the case of link failure or node failure, the system quickly selects alternative nodes according to the communication potential index and reconstructs the data transmission path. Through dynamic path reconstruction, the network can quickly recover and continue communication. When selecting alternative nodes, the nodes are sorted according to their communication potential index, and those with higher potential are preferred.

[0035] Specifically, according to the sorting result of the real-time communication potential index and the link survival prediction information, the specific process of selecting the optimal candidate node to take over the relay function is as follows: Receive the real-time communication potential index broadcast by neighboring nodes, generate a candidate node list in descending order, and combine geographical location and communication range limitations to eliminate nodes beyond the effective connection range; Based on the historical channel stability data of candidate nodes and the link survival prediction model, screen nodes with a survival probability higher than the set threshold to form a highly reliable candidate set; Send a migration request to the nodes in the highly reliable candidate set and verify their current load status to ensure that their communication potential index and load level meet the takeover conditions; Achieve seamless switching through the dual-buffer routing table mechanism. The old table maintains the existing data transmission, and the new table completes the migration and takes effect within the set time; Trigger the broadcast of local topology synchronization beacons, update the routing information of all network nodes, and quickly reconstruct the path based on the communication potential index in case of link failure.

[0036] In this implementation plan, the real-time communication potential index: Each node periodically broadcasts its communication potential index, which reflects the communication ability of the node and comprehensively considers multiple factors such as energy, channel quality, and load. Generate a candidate node list in descending order: Sort the communication potential indices of all received neighboring nodes in descending order, and preferentially select nodes with higher communication potential. Geographical location and communication range limitation: Combine the geographical location (such as distance, terrain, etc.) and the communication range (such as the preset communication radius) to eliminate nodes outside the effective connection range, ensuring that the subsequently selected nodes are physically connectable. Historical channel stability data: This part of the data records the channel stability of candidate nodes, that is, the stability degree of the communication channel between the node and other nodes, including packet loss rate, delay, signal strength fluctuation, etc. during past communication processes. Link survival prediction model: This model uses the node's historical data and the current network state to predict the survival probability of the link between the node and other nodes. By analyzing the survival probability of candidate nodes, select those nodes whose survival probability is higher than the set threshold. These nodes are considered to be able to maintain stable communication within a certain period in the future. High-reliability candidate set: The selected nodes form a high-reliability candidate set. These nodes have a higher survival probability and stronger communication ability and are suitable as relay node successors. Send a migration request to the nodes in the high-reliability candidate set and verify their current load status to ensure that their communication potential index and load level meet the replacement conditions. Migration request: Send a migration request to the nodes in the high-reliability candidate set, asking these nodes whether they are willing to take over the relay function. The migration request not only focuses on the communication potential index of the node but also needs to consider the current load situation of the node. Verify the load status: Verify the load situation of the candidate node (such as queue length, processing capacity, delay, etc.). If the node load is too high, it is not suitable to take over the relay task. Only when both the communication potential index and the load level of the node meet the replacement conditions will the node be selected. Replacement conditions: Generally speaking, the communication potential index of the candidate node needs to reach a certain threshold, and the load level should be lower than a certain set value to ensure that the node can efficiently undertake the relay task. Achieve seamless switching through the dual-buffer routing table mechanism. The old table maintains the existing data transmission, and the new table completes the migration and takes effect within the set time. Dual-buffer routing table mechanism: The dual-buffer routing table mechanism means that two routing tables are maintained during data transmission: one is the old table, which is used to maintain the current communication path, and the other is the new table, which is used to store new routing information. When the system decides to switch the relay node, the new routing table has been prepared in advance. Seamless switching: The new routing table completes the migration and takes effect within the set time, thus ensuring that the data transmission is not affected by the change of the relay node. Even during the switching process, the data flow is not interrupted. This seamless switching helps to improve the stability of the network. Trigger local topology synchronization beacon broadcasting, update the routing information of all network nodes, and quickly reconstruct the path based on the communication potential index when the link is interrupted.Local Topological Synchronization Beacon Broadcasting: When a relay node changes, the system triggers local topological synchronization beacon broadcasting, which notifies surrounding nodes to update their local routing information. This is a measure to ensure the consistency of the network topology and prevent communication failures caused by some nodes failing to learn the new topology in a timely manner. Routing Information Update for All Nodes in the Network: The broadcast beacon can help all nodes in the network update their routing tables in real time, ensuring that each node in the network can understand the current communication path. Rapid Path Reconstruction in Case of Link Interruption: In the case of a link interruption, the system will quickly select a new relay node based on the communication potential index of the nodes and reconstruct the path by updating the routing table. The selected node must have strong communication potential to ensure stable data transmission.

[0037] Specifically, according to the local channel quality, remaining energy, and the urgency of the rescue task, a multi-objective reward function is designed. The specific process of converging to the Nash equilibrium state through anti-jamming beacon interaction incremental game decision-making information in the scenarios of dynamic topology and partial node failures is as follows: According to the local channel quality, remaining energy, and the urgency of the rescue task, a channel gain term, an energy penalty term, and a task priority weighting term are designed respectively, and a multi-objective composite reward function is constructed; The local decision-making information, including the node action strategy, the reward value increment, and the link state summary, is compressed and encoded by anti-jamming beacons and broadcast to neighboring nodes using frequency hopping spread spectrum technology; Nodes update their local strategies based on the received game information, and iteratively optimize the action selection probability distribution through distributed reinforcement learning; When the strategy difference degree of all nodes in the network continuously drops below the set threshold and the key performance indicators tend to be stable, it is determined that the system converges to the Nash equilibrium state.

[0038] In this implementation plan, according to the local channel quality, remaining energy, and the urgency of the data task, a channel gain term, an energy penalty term, and a task priority weighting term are designed respectively, and a multi-objective composite reward function is constructed. Channel Gain Term: Channel gain refers to the quality of the channel between nodes. A high-quality channel will improve the reliability and efficiency of data transmission. Therefore, in the reward function, the quality of the channel has a positive impact on the node selection strategy. The formula can be expressed as: ; is the channel gain of node , and SNR is the signal-to-noise ratio. Energy Penalty Term: The remaining energy of a node affects its ability to continue to undertake communication and computing tasks. Nodes with low energy should be penalized to avoid their participation in tasks. The energy penalty term can be expressed by the following formula: ; is the energy penalty term of node , is the current remaining energy of the node, is the maximum energy of the node, is the coefficient of energy penalty. Weighted sum of task priorities: The urgency of rescue tasks determines the priorities of different tasks. Higher-priority tasks should receive more transmission resources. The weighted term of task priority can be expressed as: ; is the weighted term of task , is the urgency of task , is the weighted coefficient of task priority. Multi-objective composite reward function: The reward function designed according to the above three parts combines the channel gain, energy penalty, and weighted term of task priority through weighting to form a multi-objective reward function: ; is the comprehensive reward value when node executes task . are the weight coefficients of each item, representing the relative importance of channel gain, energy penalty, and task priority. Local decision information compression coding through anti-jamming beacon: The local decision information, including the node action strategy, reward value increment, and link state summary, is compressed and encoded by the anti-jamming beacon, and broadcast to neighboring nodes using the frequency-hopping spread-spectrum technology. Local decision information compression coding: Each node generates decision information based on the current strategy and reward value. This information includes the node's action strategy (selected action), reward value increment (reward change obtained after a specific action), and link state (current network state, such as whether the link is available, quality, etc.). Anti-jamming beacon: To resist external interference, an anti-jamming beacon technology is used to compress and encode this decision information to ensure the stable propagation of information in a complex environment. Frequency-hopping spread-spectrum technology: The frequency-hopping spread-spectrum technology avoids interference by randomly changing the signal frequency, improving the reliability of transmission. Through this technology, the compressed local decision information is broadcast to neighboring nodes, enabling nodes to share decision information within a large range. The node updates its local strategy based on the received game information, and iteratively optimizes the action selection probability distribution through distributed reinforcement learning. Game information update: After each node receives the game information from neighboring nodes, it adjusts its local strategy according to this information. Through game theory, nodes can update their strategies based on the strategies of other nodes to achieve the purpose of optimizing the overall network performance. Distributed reinforcement learning: Each node iteratively optimizes through the reinforcement learning algorithm, does not rely on centralized control, and can dynamically adjust its behavior according to the network environment. The node adjusts its strategy by observing the reward after each decision to maximize the long-term return. The update process can be expressed as: ; is the Q value when node selects action , representing the long-term return of this behavior. is the learning rate, which determines the learning speed. is the discount factor, representing the weight of future rewards. is the immediate reward when the node selects an action at that time. is the other action selected, and is the next state reached through this action. When the policy difference degree of all network nodes continues to be lower than the set threshold and the key performance indicators tend to be stable, it is determined that the system converges to the Nash equilibrium state. Policy difference degree: The policy difference degree of all network nodes is an index to measure the decision-making consistency of each node in the whole system. If the policies of most nodes tend to be the same, it indicates that each node in the system has reached a certain balance in the game. Key performance indicator stability: Key performance indicators, such as network throughput, latency, link stability, etc., reach the expected stable state, indicating that the resource allocation in the network has reached the optimal. Convergence judgment: When the policy difference degree of all network nodes in the whole network is less than the set threshold and the key performance indicators have been stable, the system is determined to have converged to the Nash equilibrium state, indicating that the behaviors among nodes have reached a stable equilibrium point.

[0039] Specifically, the specific process of generating a multi-hop scheduling strategy for collaborative optimization of energy consumption balance, transmission real-time performance and network survivability is as follows: Dynamically allocate relay tasks based on the Nash equilibrium strategy, limit the continuous working duration of high-load nodes, and trigger low-load nodes to take over part of the forwarding traffic; Reserve dedicated low time slots for key data streams such as vital signs and position coordinates, and allow them to preempt non-critical data time slots to ensure transmission real-time performance; Deploy redundant paths in the potential risk areas of the disaster heat map, and quickly activate the optimal backup link based on the difference of the communication potential index when the main path is interrupted; Periodically trigger the synchronization of the whole network topology state, dynamically update the node routing table and suppress the spread of local congestion to ensure network survivability.

[0040] In this implementation plan, relay tasks are dynamically allocated based on the Nash equilibrium strategy, the continuous working duration of high-load nodes is limited, and low-load nodes are triggered to take over part of the forwarding traffic. Dynamically allocate relay tasks based on the Nash equilibrium strategy: In the network, each node allocates relay tasks through the Nash equilibrium strategy in game theory according to the current load status and task requirements. Specifically, the node dynamically selects whether to participate in the relay task according to the current network state (such as channel quality, node energy, load, etc.). If the load of the node is high, it will avoid carrying the relay task for a long time. Limit the continuous working duration of high-load nodes: In order to prevent the performance of some nodes from degrading due to overload, it is necessary to set a working time limit for high-load nodes to prevent them from overworking and thus affecting the overall network performance. By setting the upper limit of the node working time, the burden on the node can be effectively reduced. The formula is expressed as: ; is the current working duration of node i, is the maximum working duration of node i, which prevents high - load nodes from timing out. Trigger low - load nodes to take over part of the forwarding traffic: When the load of a node is below a certain threshold, it will automatically take over part of the forwarding traffic of high - load nodes. The amount of traffic taken over can be expressed by the following formula: ; is the forwarding traffic taken over by node i. is the current load of node i, is the maximum load capacity of the node, is the adjustment coefficient, is the total traffic. Reserve dedicated low - time - slot resources for key data streams of vital signs and position coordinates, allowing them to preempt non - critical data time slots to ensure transmission real - time. Dedicated low - time - slot resources: To ensure the real - time transmission of key data (such as vital signs and position coordinate data), the system reserves dedicated low - time - slot resources for these data streams. These low - time - slot resources ensure the priority of key data streams and prevent transmission delays or losses. Preempt non - critical data time slots: When the transmission requirements of key data streams are small, the system allows them to preempt the time slots of non - critical data to ensure real - time. This is achieved through a priority scheduling mechanism. The formula is expressed as: ; is the time slot reserved by node i for key data streams, is the priority of key data, is the time - slot resources allocated to node i. Deploy redundant paths in potential risk areas of the disaster heat map. When the main path is interrupted, quickly activate the optimal backup link based on the difference in communication potential index Redundant path deployment: In high - risk disaster areas, to ensure the reliability and survivability of communication, redundant paths need to be deployed. Partial deployment of redundant paths can be based on the disaster heat map, and areas with a lower probability of disaster occurrence and stronger communication potential are selected for redundant link deployment. Activate the backup link based on the difference in communication potential index: When the main path is interrupted, the system quickly activates potential backup links according to the communication potential index (CPI). The difference in communication potential index can be expressed by the following formula: ; is the communication potential index of the main path, is the communication potential index of the backup path, It is the difference in potential indices. If the difference is greater than the set threshold, the system will activate the backup link. Periodically trigger the synchronization of the whole network topology state, dynamically update the node routing table and suppress the spread of local congestion to ensure the network survivability. Whole network topology state synchronization: To ensure the dynamic stability of the network, nodes will periodically trigger the topology state synchronization to ensure that the routing tables and network states of all nodes are consistent. During the synchronization process, nodes will broadcast their routing information and current status to ensure that all nodes in the network have a consistent understanding of the current network situation. Dynamically update the node routing table: Based on the topology synchronization information, nodes will update the local routing table and adjust the path of the data flow to cope with the current network state and topology changes. Suppress the spread of local congestion: When network congestion occurs in a local area, the system avoids the spread of congestion to the whole network by dynamically adjusting the data flow path. By selecting the optimal path and restricting the traffic at the congested nodes, the impact of local congestion can be effectively suppressed. The formula is expressed as: ; is the congestion level of node , is the set of nodes adjacent to node , is the traffic of the neighboring node .

[0041] Please refer to Figure 2, A multi-hop communication node intelligent scheduling system, comprising the following modules: a communication potential analysis module, a relay node selection module, a role migration trigger module, and an autonomous game decision-making module; the communication potential analysis module is used to collect in real time the remaining energy, channel fading parameters, and link load data of the UAV and ground emergency terminal communication nodes, calculate the communication potential index through dynamic weight fusion, and is used to comprehensively quantify the multi-hop relay ability of the nodes under energy constraint and channel mutation; the relay node selection module is used to set relay node selection thresholds for different disaster types according to the communication potential index, preferentially select communication nodes with a communication potential index higher than the disaster situation adaptation threshold as relay nodes, construct an initial hybrid topology with priority for anti-destruction, and perform load pre-allocation on key rescue data streams in combination with the disaster heat distribution map; the role migration trigger module is used to trigger a role migration operation when it detects that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to a sudden drop in energy or channel disturbance, broadcast a migration request carrying an emergency level identifier to neighboring nodes, and select the optimal candidate node to take over the relay function according to the real-time communication potential index sorting result and link survival prediction information; the autonomous game decision-making module is used to model each node as an autonomous game agent, whose action space includes four states: relay, terminal, standby, and emergency communication. A multi-objective reward function is designed according to the local channel quality, remaining energy, and urgency of the rescue task. Through anti-interference beacon interaction and incremental game decision-making information, it converges to the Nash equilibrium state in the scenarios of dynamic topology and partial node failure, and generates a multi-hop scheduling strategy for collaborative optimization of energy consumption balance, transmission real-time performance, and network anti-destruction ability.

[0042] In this implementation solution, there is a communication potential analysis module: This module calculates the communication potential index by dynamically collecting and integrating multi-dimensional data such as the remaining energy, channel fading parameters, and link load of drones and ground emergency terminal nodes. This method is different from traditional single evaluation methods and can dynamically calculate the communication potential index by integrating multiple factors (such as energy, channel quality, load, etc.), thus more accurately reflecting the multi-hop relay ability of nodes in different network environments. Especially in the case of energy limitation and channel mutation, it can ensure the stability of the system communication performance. There is a relay node selection module: This module selects relay nodes based on the communication potential index and pre-distributes the load for key rescue data streams in combination with the disaster heat distribution map. This strategy can set the relay node selection threshold according to different disaster types, preferentially select nodes with high potential index as relays, and construct a hybrid topology structure with priority on anti-destruction. This not only improves the reliability of the network but also can dynamically optimize resource allocation according to different situations of disasters to ensure the efficient transmission of rescue data streams. There is a role migration trigger module: This module designs a role migration mechanism, which triggers the migration operation by real-time monitoring the communication potential index of nodes and when it is lower than the safety threshold. The migration of nodes selects the optimal candidate nodes by real-time sorting the communication potential index and link survival prediction information. This role migration mechanism based on dynamic monitoring and prediction can ensure seamless switching of relay tasks when nodes fail or are overloaded, avoiding the communication interruption problem caused by the failure of relay nodes in traditional methods. There is an autonomous game decision-making module: This module adopts the Nash equilibrium model in game theory, regards each communication node as an autonomous game agent, designs a multi-objective reward function to balance objectives such as energy consumption, transmission real-time performance, and anti-destruction. Nodes optimize their own decisions through distributed reinforcement learning and interact with incremental game decision-making information through anti-interference beacons. In the case of dynamic topology and partial node failures, the system can adaptively converge to the Nash equilibrium state, thus realizing the collaborative optimization of the objectives of energy consumption balance and network anti-destruction.

[0043] In summary, this application has at least the following effects: A multi-hop communication node intelligent scheduling method and system. Through the multi-dimensional data fusion and dynamic weight calculation of the communication potential analysis module, the system can evaluate and optimize the multi-hop relay ability of nodes in real time, ensuring that the communication link remains stable and reliable even in the face of channel mutations and energy limitations in a disaster environment. The selection process of relay nodes takes into account factors such as disaster type, node status, and task urgency, and can be dynamically adjusted according to the real-time communication potential index to ensure the selection of the most suitable node for relay, thereby improving the network's survivability and adaptability. When the communication potential index of a node decreases or a link problem occurs, the role migration trigger module can respond quickly and migrate the relay task to a neighboring node with stronger communication potential in real time, avoiding communication interruption caused by the failure of a single node and enhancing the system's self-healing ability and fault tolerance. The autonomous game decision-making module combines game theory models with distributed reinforcement learning to dynamically optimize the behavior decisions of nodes, synergistically optimizing multiple objectives such as energy consumption, transmission real-time performance, and network survivability, ensuring that the system can adaptively adjust in a dynamic environment and improving the overall network efficiency. By combining the reserved resource allocation for critical rescue data streams with the disaster heat map, the system can ensure that the priority of critical task data streams (such as vital signs and location coordinate data) is guaranteed during transmission, improving the transmission real-time performance and success rate of critical data. The system can optimize the network state and load by updating routing information and adjusting resource allocation in real time in scenarios with dynamic topologies and partial node failures, ensuring efficient and stable operation in a disaster situation.

[0044] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0045] The present invention is described with reference to the flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, as well as the combination of flows and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices produce a device for realizing the functions specified in Figure 1 one or more of the flows or multiple flows and / or blocks Figure 1 one or more of the blocks or multiple blocks.

[0046] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means that implement the function specified in one or more processes and / or blocks Figure 1 in one or more processes and / or blocks Figure 1 specified in one or more blocks or blocks.

[0047] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable apparatus provide steps for implementing the function specified in one or more processes and / or blocks Figure 1 in one or more processes and / or blocks Figure 1 specified in one or more blocks or blocks.

[0048] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn of the basic inventive concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications that fall within the scope of the present invention.

[0049] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. A multi-hop communication node intelligent scheduling method, characterized in that: The following steps are involved: S1. Collect the residual energy, channel fading parameters and link load data of the communication nodes of the UAV and the ground emergency terminal in real time, calculate the communication potential index through dynamic weight fusion, and use it to comprehensively quantify the multi-hop relay capability of the node under energy limitation and channel mutation; S2. According to the communication potential index, set the relay node selection threshold for different disaster types, give priority to selecting communication nodes with a communication potential index higher than the disaster adaptation threshold as relay nodes, build an initial hybrid topology structure with priority on anti-destruction, and pre-allocate the load of key rescue data streams in combination with the disaster thermal distribution map; S3. When it is detected that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to a sudden drop in energy or channel disturbance, the role migration operation is triggered, and a migration request carrying an emergency level identifier is broadcast to neighboring nodes. Based on the real-time communication potential index sorting results and link survival prediction information, the best candidate node is selected to take over the relay function; S4. Each node is modeled as an autonomous game agent, whose action space includes four states: relay, terminal, standby and emergency communication. A multi-objective reward function is designed according to the local channel quality, remaining energy and the urgency of the rescue mission. The incremental game decision information is interactively obtained through anti-interference beacons. It converges to the Nash equilibrium state under dynamic topology and partial node failure scenarios, and generates a multi-hop scheduling strategy that coordinates the optimization of energy consumption balance, transmission real-time and network anti-destruction.

2. A multi-hop communication node intelligent scheduling method according to claim 1, characterized in that: The specific process of real-time collection of residual energy, channel fading parameters and link load data of UAV and ground emergency terminal communication nodes is as follows: The energy monitoring modules deployed on drones and ground terminals periodically collect the remaining energy value of the battery and predict the energy decay trend in combination with the charge and discharge rate. Through the real-time communication feedback mechanism of the link layer, the received signal strength and bit error rate between each pair of communication nodes are counted, the instantaneous channel fading level is calculated, and the fluctuation amplitude in a short period of time is analyzed by sliding, and the channel quality change gradient is extracted to determine whether there is a risk of sudden change in the channel; Combining the forwarding queue status and actual load conditions of each node, the cache queue length, average waiting delay and packet queuing frequency are monitored in real time to form a load vector that reflects the link load level. The load peaks in different time slices are normalized to generate standardized link pressure characteristic values.

3. A multi-hop communication node intelligent scheduling method according to claim 2, characterized in that: The specific process of calculating the communication potential index through dynamic weight fusion is as follows: For each communication node, the standardized index values ​​are extracted from the three dimensions of residual energy, channel fading and link load, and their numerical ranges are unified; According to the current disaster type, node role task and task urgency parameters, differentiated dynamic weights are assigned to the three indicators to generate a weight coefficient group that reflects the current scheduling priority requirements; The three weighted indicators are fused to construct a three-dimensional feature vector, and the communication potential index representing the multi-hop relay capability of the current node is generated by calculating its mapping function value in a specific scheduling strategy space.

4. A multi-hop communication node intelligent scheduling method according to claim 3, characterized in that: According to the communication potential index, the specific process of setting the relay node selection threshold for different disaster types is as follows: Obtain the communication guarantee level requirements corresponding to the current disaster type, and extract the minimum value of the communication potential index that successfully maintains the stability of the communication link in various disaster scenarios as the reference threshold based on the distribution range of the communication potential index in historical disaster emergency communication tasks; Combined with the current task urgency and node distribution density, a weighted correction coefficient is set to dynamically adjust the basic threshold to form a disaster-adaptive communication potential selection threshold; Sort the communication potential indexes of all candidate nodes and select nodes with a value higher than the set threshold as the candidate set of relay nodes; Based on the candidate node set, the target node suitable for relay is further selected based on the following specific criteria: Geographic location constraints: Eliminate nodes that are located in communication blind spots, are severely blocked by terrain, or are farther away from the target service area than the preset communication radius; Link connectivity constraint: select nodes whose number of adjacent nodes is higher than the average connectivity threshold through adjacency matrix calculation; Energy distribution balance constraint: Calculate the energy variance of the candidate node set, give priority to nodes with medium to high power levels, and eliminate nodes whose current remaining energy is lower than a certain percentage of the average value of the entire network.

5. A multi-hop communication node intelligent scheduling method according to claim 4, characterized in that: The specific process of constructing an initial hybrid topology structure with survivability priority and pre-allocating the load of key rescue data streams in combination with the disaster thermal distribution map is as follows: The backbone nodes are selected by the communication potential index, some of which are used as cluster head nodes to cover high-priority rescue areas, and the remaining backbone nodes build a mesh redundancy layer to form a communication topology with multi-path connection capabilities; Rescue data is managed in different levels, and different transmission resources and path redundancy levels are allocated according to the importance of the data. High-priority data is configured with fixed time slots and multi-path transmission strategies, while low-priority data uses dynamic resource scheduling. Backup channels are preset for critical data to enhance link reliability. Redundant cluster head nodes are deployed in potential high-risk areas to form a cross-coverage structure, and a load sensing mechanism is set up at key nodes to dynamically migrate part of the traffic to adjacent nodes according to the current load status of the node to achieve link load balancing; The topology structure periodically broadcasts topology status beacons to drive each node to update local routing information in real time, and when a link is interrupted, it quickly selects alternative nodes to complete path reconstruction based on communication potential indicators.

6. A multi-hop communication node intelligent scheduling method according to claim 5, characterized in that: According to the ranking results of the real-time communication potential index and the link survival prediction information, the specific process of selecting the best candidate node to take over the relay function is as follows: Receive the real-time communication potential index broadcast by neighboring nodes, generate a list of candidate nodes in descending order, and remove nodes that are beyond the effective connection range by combining geographical location and communication range restrictions; Based on the historical channel stability data of candidate nodes and the link survival prediction model, nodes with a survival probability higher than the set threshold are screened to form a highly reliable candidate set; Send migration requests to nodes in the high-reliability candidate set and verify their current load status; After the migration verification is passed, the local topology synchronization beacon broadcast is triggered to update the routing table of all network nodes. If the link is interrupted, the network path is quickly reconstructed according to the communication potential index and topology status to ensure uninterrupted data transmission.

7. A multi-hop communication node intelligent scheduling method according to claim 6, characterized in that: A multi-objective reward function is designed based on the local channel quality, remaining energy, and the urgency of the rescue mission. The specific process of converging to the Nash equilibrium state in the dynamic topology and partial node failure scenarios is as follows: According to the local channel quality, remaining energy and rescue mission urgency, the channel gain term, energy penalty term and task priority weighted term are designed respectively to construct a multi-objective composite reward function. The local decision information, including node action strategy, reward value increment and link status summary, is compressed and encoded by anti-interference beacon, and broadcasted to neighboring nodes using frequency hopping spread spectrum technology. The node updates the local strategy based on the received game information and iteratively optimizes the probability distribution of action selection through distributed reinforcement learning; When the strategy difference of nodes in the entire network continues to be lower than the set threshold and the key performance indicators tend to be stable, it is determined that the system converges to the Nash equilibrium state.

8. A multi-hop communication node intelligent scheduling method according to claim 7, characterized in that: The specific process of generating a multi-hop scheduling strategy that coordinates energy consumption balance, transmission real-time performance, and network invulnerability optimization is as follows: Dynamically allocate relay tasks based on the Nash equilibrium strategy, limit the continuous working time of high-load nodes, and trigger low-load nodes to take over part of the forwarding traffic; Reserve dedicated low-slot resources for preset critical data streams, allowing them to preempt non-critical data slots to ensure real-time transmission; Deploy redundant paths in potential risk areas of the disaster heat map, and quickly activate the optimal backup link based on the difference in communication potential index when the main path is interrupted; Periodically trigger the synchronization of the entire network topology status, dynamically update the node routing table and suppress the spread of local congestion to ensure the network's anti-destruction capabilities.

9. A multi-hop communication node intelligent scheduling system, applied to a multi-hop communication node intelligent scheduling method according to any one of claims 1 to 8, characterized in that: It includes the following modules: communication potential analysis module, relay node selection module, role migration trigger module, and autonomous game decision module; The communication potential analysis module is used to collect the residual energy, channel fading parameters and link load data of the communication nodes of the UAV and the ground emergency terminal in real time, and calculate the communication potential index through dynamic weight fusion, which is used to comprehensively quantify the multi-hop relay capability of the node under energy limitation and channel mutation; The relay node selection module is used to set the relay node selection threshold for different disaster types according to the communication potential index, give priority to selecting communication nodes with a communication potential index higher than the disaster adaptation threshold as relay nodes, build an initial hybrid topology structure with priority on anti-destruction, and pre-distribute the load of key rescue data streams in combination with the disaster thermal distribution map; The role migration trigger module is used to trigger the role migration operation when it is detected that the communication potential index of the relay node is lower than the dynamically updated safety threshold due to energy drop or channel disturbance, broadcast a migration request carrying an emergency level identifier to the neighboring nodes, and select the best candidate node to take over the relay function according to the real-time communication potential index sorting result and link survival prediction information; The autonomous game decision module is used to model each node as an autonomous game agent, whose action space includes four states: relay, terminal, standby and emergency communication. A multi-objective reward function is designed according to the local channel quality, remaining energy and the urgency of the rescue mission. Through the interactive incremental game decision information of anti-interference beacons, it converges to the Nash equilibrium state under dynamic topology and partial node failure scenarios, and generates a multi-hop scheduling strategy with coordinated optimization of energy consumption balance, transmission real-time and network anti-destruction.

Images