A Method and System for Optimizing Downhole Communication Links Based on Self-Organizing Networks

Abstract

This invention provides a method and system for optimizing downhole communication links based on self-organizing networks, relating to the field of communication link technology. The method includes acquiring communication node information, constructing a network topology, and determining network characteristic parameters; constructing a link quality prediction model to evaluate link quality; performing dynamic network partitioning and establishing a node priority ranking table to select the optimal transmission path; and real-time monitoring of link status, triggering node self-organizing reconfiguration when necessary. This invention can improve the connectivity, stability, and transmission efficiency of downhole communication networks, reduce communication latency, and enhance the safety of downhole operations.

Description

Technical Field

[0001] This invention relates to communication link technology, and more particularly to a method and system for optimizing downhole communication links based on self-organizing networks. Background Technology

[0002] Underground communication networks, as crucial infrastructure for underground engineering projects such as mining and tunnel construction, play a key role in ensuring worker safety and improving production efficiency. With the development of intelligent mining technology, the number of devices and sensors in the underground environment has increased dramatically, placing higher demands on the stability and reliability of communication networks. Traditional underground communication systems mainly use wired communication methods, such as fiber optics and leaky cables. However, in the complex and ever-changing underground environment, these fixed communication facilities are easily damaged. Wireless communication technology, due to its flexibility and ease of deployment, has gradually been introduced into the underground communication field. By deploying multiple wireless communication nodes to build a self-organizing network, reliable data transmission can be achieved.

[0003] The underground environment is complex and ever-changing, with intricate tunnel structures and numerous obstacles such as rock walls and support equipment. Signal propagation is severely attenuated and affected by multipath effects, leading to unstable communication link quality. During underground operations, the mining face advances continuously, requiring frequent adjustments to communication node positions and dynamic changes in the network topology. Traditional static routing algorithms struggle to adapt to this highly dynamic environment. Furthermore, the service requirements vary significantly across different areas underground. For example, applications such as video surveillance, equipment control, and environmental monitoring have varying bandwidth and latency requirements. However, existing communication systems lack intelligent network resource allocation mechanisms and cannot flexibly optimize transmission paths based on service load characteristics, resulting in low network resource utilization efficiency.

[0004] To address these issues, there is an urgent need for an underground communication link optimization method that can sense network status, predict link quality, and dynamically adjust communication paths to improve the stability, reliability, and resource utilization efficiency of underground communication networks and meet the high-quality communication support requirements of smart mine construction. Summary of the Invention

[0005] This invention provides a method and system for optimizing downhole communication links based on self-organizing networks, which can solve the problems in the prior art.

[0006] A first aspect of this invention provides a method for optimizing downhole communication links based on self-organizing networks, comprising:

[0007] Acquire location information, channel status information, and service load information of multiple communication nodes distributed underground;

[0008] Based on the location information and the channel state information, a network topology structure reflecting the connectivity between nodes is constructed through a topology discovery mechanism. The connectivity, node distribution density and topology stability of the network topology structure are analyzed to determine network characteristic parameters.

[0009] Based on the network characteristic parameters, a link quality prediction model integrating network state awareness is constructed. The link quality model is used to determine the link quality level between each communication node, and a predicted link quality assessment result is formed.

[0010] Combining the predicted link quality assessment results with the service load information, the communication network is dynamically partitioned, the transmission nodes in each partition are determined, and a node priority ranking table is established. The optimal transmission path is selected according to the node priority ranking table. The link status of the optimal transmission path is periodically probed. When a reduction in link transmission quality is detected and an adjustment threshold is triggered, the node self-organizing reconstruction process is activated and the topology discovery mechanism is re-executed to update the network characteristic parameters.

[0011] The target path is recalculated based on the updated network feature parameters, and a communication link is established to perform data transmission.

[0012] Based on the location information and the channel state information, a network topology reflecting the connectivity relationships between nodes is constructed through a topology discovery mechanism. The network topology's connectivity, node density, and topology stability are analyzed to determine network characteristic parameters, including:

[0013] Based on location information, the spatial distance between each communication node is calculated. Combined with the signal strength and channel fading characteristics in the channel state information, the reachability relationship between each communication node is determined through the topology discovery mechanism, and a network topology structure containing node identifiers and inter-node connection relationships is constructed.

[0014] During the execution of the topology discovery mechanism, neighbor discovery response information of each communication node is collected and the number of neighbor nodes is counted. The connectivity degree, which reflects the overall connectivity of the network, is calculated by combining the number of connected subgraphs in the network.

[0015] Based on the location information, the underground space is divided into multiple regional units. The distribution of communication nodes in each regional unit is statistically analyzed. Based on the variance analysis of the regional node distribution, the node distribution density reflecting the uniformity of spatial distribution is determined.

[0016] By periodically executing the topology discovery mechanism, the changes in the connection relationships between nodes within a preset time window are recorded, and the topology stability, which reflects the degree of dynamic change in the network topology, is calculated based on the frequency of changes in the connection relationships.

[0017] The connectivity, node distribution density, and topological stability are subjected to hierarchical weighted mapping, and dynamic adjustment coefficients are set according to the importance of each parameter to form network feature parameters.

[0018] Based on location information, the spatial distance between communication nodes is calculated. Combined with signal strength and channel fading characteristics from channel state information, a topology discovery mechanism is used to determine the reachability relationships between communication nodes. This constructs a network topology structure that includes node identifiers and inter-node connectivity relationships, including:

[0019] The spatial distance between each communication node is calculated based on the location information. Based on the spatial distance and the geometric constraint relationship between the underground roadway, communication node pairs with straight propagation paths and communication node pairs with diffraction propagation paths are identified to obtain propagation path type identifiers.

[0020] The topology discovery mechanism is used to calculate the channel fading characteristics of the straight propagation path and the diffraction propagation path based on the signal strength in the channel state information and the propagation path type identifier. The straight propagation path is calculated based on the spatial distance attenuation characteristics, and the diffraction propagation path is calculated based on the multipath superposition effect.

[0021] The signal strength and the channel fading characteristics are comprehensively evaluated. When the comprehensive evaluation result meets the communication quality constraints, it is determined that the corresponding communication node pairs have a reachability relationship, and connection weights are assigned to the communication node pairs with a reachability relationship.

[0022] Based on the topology discovery mechanism, a network topology structure containing node identifiers, inter-node connection relationships, and connection weights is constructed according to the reachability relationship and the connection weight.

[0023] Based on the network characteristic parameters, a link quality prediction model integrating network state awareness is constructed. This model is then used to determine the link quality level between each communication node, resulting in a predicted link quality assessment result, including:

[0024] Connectivity, node distribution density, and topology stability are extracted from network feature parameters, and channel state information and service load information between communication nodes are obtained within a historical time period. Based on the temporal correlation between connectivity, node distribution density, topology stability, channel state information, and service load information, a link quality prediction model with fused network state awareness is constructed.

[0025] By inputting the current network characteristic parameters, channel state information, and traffic load information into the link quality prediction model, the predicted link quality values ​​between communication nodes in the future time period are obtained.

[0026] Based on the predicted link quality values ​​and in conjunction with a preset link quality grading standard, the link quality levels between communication nodes are divided into multiple quality levels, and a quality weight coefficient is assigned to each quality level. The quality weight coefficient reflects the degree of influence of different quality levels on path selection.

[0027] The predicted link quality value is combined with the quality level to form the predicted link quality assessment result.

[0028] Combining the predicted link quality assessment results with the service load information, the communication network is dynamically partitioned, the transmission nodes within each partition are determined, and a node priority ranking table is established. The optimal transmission path is selected based on the node priority ranking table, including:

[0029] Based on the predicted link quality values ​​in the predicted link quality assessment results, communication node connections in the network topology with predicted link quality values ​​lower than a preset quality threshold are identified. In-depth analysis of communication node connections is performed in conjunction with the link awareness mechanism, and the analysis results are used as partition boundary markers to dynamically partition the communication network.

[0030] Based on the link-aware mechanism, the status of each network partition is continuously monitored, the service load information and location information of each communication node are extracted, and the communication nodes with loads below the load threshold are identified as candidate transmission nodes for the network partition through the resource scheduling algorithm.

[0031] For candidate transmission nodes in each network partition, based on the resource scheduling algorithm and the predicted link quality assessment results and service load information, calculate the comprehensive path quality score and remaining transmission capacity of the node, and establish a node priority ranking table based on the score results;

[0032] Based on the node priority ranking table, the transmission node with the highest ranking and that meets the monitoring requirements of the link awareness mechanism is selected as the relay node to construct the optimal transmission path from the source node to the data aggregation node.

[0033] Based on the resource scheduling algorithm, combined with the predicted link quality assessment results and service load information, the comprehensive path quality score and remaining transmission capacity of the nodes are calculated. A node priority ranking table is then established based on the score results, including:

[0034] Extract the predicted link quality values ​​from each communication node to the data aggregation node from the predicted link quality assessment results. Based on the resource scheduling algorithm, the predicted link quality values ​​and the path transmission hop count are weighted and fused to calculate the comprehensive path quality score from each communication node to the data aggregation node.

[0035] Obtain the service load information of each communication node, wherein the service load information includes the current cache queue length of the node and the maximum cache capacity of the node;

[0036] The resource scheduling algorithm determines the remaining transmission capacity of each communication node based on the ratio of the current cache queue length of the node to the maximum cache capacity of the node, and multiplies the remaining transmission capacity as a load balancing factor with the comprehensive path quality score to obtain a comprehensive node score that takes into account both path quality and node load.

[0037] Based on the numerical value of the node's comprehensive score, the communication nodes are arranged in descending order to form a node priority ranking table that reflects the node's transmission priority level.

[0038] The process of recalculating the target path based on the updated network feature parameters and establishing a communication link to perform data transmission includes:

[0039] The updated network feature parameters are obtained, and the network status change trend of each communication node in the network feature parameters is analyzed. Based on the network status change trend, the weight coefficient of each communication node in the target path selection is dynamically adjusted to construct a path optimization strategy that reflects the dynamic adaptability of the network. Based on the path optimization strategy, the target path is determined, and a communication link is established to execute data transmission.

[0040] A second aspect of the present invention provides a downhole communication link optimization system based on a self-organizing network, comprising:

[0041] The first unit is used to acquire the location information, channel status information, and service load information of multiple communication nodes distributed underground.

[0042] The second unit is used to construct a network topology structure reflecting the connectivity between nodes based on the location information and the channel state information through a topology discovery mechanism, and to analyze the connectivity, node distribution density and topology stability of the network topology structure to determine network characteristic parameters.

[0043] The third unit is used to construct a link quality prediction model that integrates network state awareness based on the network feature parameters, and to use the link quality model to determine the link quality level between each communication node, thereby forming a predicted link quality assessment result.

[0044] The fourth unit is used to combine the predicted link quality assessment results with the service load information to dynamically partition the communication network, determine the transmission nodes in each partition and establish a node priority ranking table, select the optimal transmission path according to the node priority ranking table, periodically detect the link status of the optimal transmission path, and when a reduction in link transmission quality is detected and an adjustment threshold is triggered, activate the node self-organizing reconstruction process and re-execute the topology discovery mechanism to update the network characteristic parameters.

[0045] The fifth unit is used to recalculate the target path based on the updated network feature parameters and establish a communication link to perform data transmission.

[0046] A third aspect of the present invention provides an electronic device, comprising:

[0047] processor;

[0048] Memory used to store processor-executable instructions;

[0049] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0050] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0051] The beneficial effects of this application are as follows:

[0052] By acquiring the location information, channel status information, and service load information of underground communication nodes, and combining this with a topology discovery mechanism to construct a network topology, we can comprehensively understand the network connectivity status in complex underground environments and provide accurate basic data for link optimization.

[0053] The innovative analysis of network topology connectivity, node distribution density, and topological stability determines network characteristic parameters, enabling precise characterization of downhole communication network features and improving the accuracy of subsequent link quality prediction.

[0054] By constructing a link quality prediction model that integrates network state awareness, the system can identify potential link quality problems in advance, thereby achieving preventive optimization and avoiding communication interruptions caused by post-event remediation in traditional methods.

[0055] Dynamic partitioning is performed based on predicted link quality assessment results and service load information, and a node priority ranking table is established. This enables reasonable allocation of resources and intelligent selection of transmission paths, effectively improving network transmission efficiency and resource utilization.

[0056] A link status periodic detection mechanism was designed. When the link quality deteriorates and triggers the adjustment threshold, the node self-organizing reconstruction process can be automatically activated to realize the self-repair and optimization of the communication network, which significantly improves the stability and reliability of the downhole communication system. Attached Figure Description

[0057] Figure 1 This is a flowchart illustrating the downhole communication link optimization method based on self-organizing networks according to an embodiment of the present invention.

[0058] Figure 2This is a flowchart illustrating the preferred transmission process for network partition nodes based on low-quality link analysis, as described in an embodiment of the present invention. Detailed Implementation

[0059] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0060] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0061] Figure 1 This is a flowchart illustrating the downhole communication link optimization method based on self-organizing networks according to an embodiment of the present invention. Figure 1 As shown, the method includes:

[0062] Based on the location information and the channel state information, a network topology structure reflecting the connectivity between nodes is constructed through a topology discovery mechanism. The connectivity, node distribution density and topology stability of the network topology structure are analyzed to determine network characteristic parameters.

[0063] Based on the network characteristic parameters, a link quality prediction model integrating network state awareness is constructed. The link quality model is used to determine the link quality level between each communication node, and a predicted link quality assessment result is formed.

[0064] Combining the predicted link quality assessment results with the service load information, the communication network is dynamically partitioned, the transmission nodes in each partition are determined, and a node priority ranking table is established. The optimal transmission path is selected according to the node priority ranking table. The link status of the optimal transmission path is periodically probed. When a reduction in link transmission quality is detected and an adjustment threshold is triggered, the node self-organizing reconstruction process is activated and the topology discovery mechanism is re-executed to update the network characteristic parameters.

[0065] The target path is recalculated based on the updated network feature parameters, and a communication link is established to perform data transmission.

[0066] In one optional implementation, based on the location information and the channel state information, a network topology reflecting the connectivity between nodes is constructed through a topology discovery mechanism. The network topology's connectivity, node density, and topology stability are analyzed to determine network characteristic parameters, including:

[0067] Based on location information, the spatial distance between each communication node is calculated. Combined with the signal strength and channel fading characteristics in the channel state information, the reachability relationship between each communication node is determined through the topology discovery mechanism, and a network topology structure containing node identifiers and inter-node connection relationships is constructed.

[0068] During the execution of the topology discovery mechanism, neighbor discovery response information of each communication node is collected and the number of neighbor nodes is counted. The connectivity degree, which reflects the overall connectivity of the network, is calculated by combining the number of connected subgraphs in the network.

[0069] Based on the location information, the underground space is divided into multiple regional units. The distribution of communication nodes in each regional unit is statistically analyzed. Based on the variance analysis of the regional node distribution, the node distribution density reflecting the uniformity of spatial distribution is determined.

[0070] By periodically executing the topology discovery mechanism, the changes in the connection relationships between nodes within a preset time window are recorded, and the topology stability, which reflects the degree of dynamic change in the network topology, is calculated based on the frequency of changes in the connection relationships.

[0071] The connectivity, node distribution density, and topological stability are subjected to hierarchical weighted mapping, and dynamic adjustment coefficients are set according to the importance of each parameter to form network feature parameters.

[0072] Calculate the Euclidean distance between each communication node. For any two nodes i and j, the spatial distance can be expressed as the distance between their coordinates. Combine the measured signal strength data and apply a logarithmic distance path loss model to evaluate the signal fading characteristics. When the signal strength between two nodes exceeds a preset receiver sensitivity threshold (usually -85dBm) and the bit error rate is lower than an acceptable threshold (e.g., 10^-3), a direct communication link is determined to exist between the two nodes.

[0073] During the construction process, each node broadcasts a probe frame containing its own identifier, and neighboring nodes that receive the probe frame send response acknowledgments. By collecting these response acknowledgments, each node generates a neighbor table containing a list of nodes with whom it can communicate directly. The neighbor table records the identifiers, link quality metrics, and update timestamps of neighboring nodes. Finally, the neighbor table information of all nodes is aggregated to construct a complete network topology graph, which consists of a set of nodes V and a set of edges E, where edges represent valid communication links between nodes.

[0074] During topology discovery, network connectivity is calculated by analyzing neighbor response information. For each node i, the number ni of its directly connected neighbor nodes is counted, and the average node degree is calculated. Simultaneously, a depth-first search algorithm is used to identify the number k of connected subgraphs in the network. Network connectivity C can be expressed as the ratio of the average node degree to the ideal fully connected state, weighted by the reciprocal of the number of connected subgraphs, thus reflecting the overall connectivity of the network. When C is close to 1, it indicates a highly connected network; when C is low, it indicates poor network connectivity.

[0075] Based on the acquired location information, the underground space is divided into multiple regional units according to the actual environmental characteristics. In the mine roadway environment, spatial units can be divided according to a specification of 20 meters × 20 meters × 5 meters (length × width × height). The number of nodes in each regional unit is counted to obtain a spatial density matrix representing the node distribution. The standard deviation of the number of nodes in each regional unit is calculated to obtain the node distribution density parameter D. A low D value indicates that the node distribution is uniform; a high D value indicates that the nodes are concentrated in a specific area, and the spatial coverage is uneven.

[0076] A topology discovery mechanism is periodically executed (e.g., every 30 seconds) to record topology changes within a preset time window (e.g., 5 minutes). For each link, the number of its state changes (establishment or disconnection) is recorded. A weighting function is set to give higher weight to recent changes, and the topology change rate is calculated. The topology stability parameter S is defined as the normalized complement of the topology change rate. The closer the S value is to 1, the more stable the network topology and the fewer changes; the closer the S value is to 0, the more frequent the network topology changes.

[0077] Connectivity (C), node density (D), and topological stability (S) are weighted and fused hierarchically. Dynamic adjustment coefficients α, β, and γ are set according to the network application scenario requirements, representing the importance of connectivity, density, and stability in the network characteristics, respectively. For example, in monitoring applications requiring high reliability, α=0.5, β=0.2, and γ=0.3 can be set; while in communication applications requiring wide coverage, α=0.3, β=0.5, and γ=0.2 can be set. The final network characteristic parameter F is obtained through weighted combination calculation, which comprehensively reflects the overall characteristics of the network topology.

[0078] In an application example, 50 communication nodes were deployed in a mining area of ​​a metal mine. The network topology was constructed using the method described above, and network characteristics were analyzed. The calculated connectivity C=0.78, indicating good network connectivity; the node density D=0.65, indicating a relatively balanced node distribution with some densely populated areas; and the topology stability S=0.82, indicating a relatively stable network structure. With weights α=0.4, β=0.3, and γ=0.3, the final calculated network characteristic parameter F=0.756, which provides an important basis for subsequent routing and network optimization.

[0079] By constructing and analyzing network topology in this way, we can accurately grasp the characteristics of underground communication networks and provide a scientific basis for communication quality assurance and network optimization.

[0080] In one optional implementation, the spatial distance between each communication node is calculated based on location information, and the reachability relationship between each communication node is determined through a topology discovery mechanism by combining the signal strength and channel fading characteristics in the channel state information, thereby constructing a network topology structure that includes node identifiers and inter-node connectivity relationships.

[0081] The spatial distance between each communication node is calculated based on the location information. Based on the spatial distance and the geometric constraint relationship between the underground roadway, communication node pairs with straight propagation paths and communication node pairs with diffraction propagation paths are identified to obtain propagation path type identifiers.

[0082] The topology discovery mechanism is used to calculate the channel fading characteristics of the straight propagation path and the diffraction propagation path based on the signal strength in the channel state information and the propagation path type identifier. The straight propagation path is calculated based on the spatial distance attenuation characteristics, and the diffraction propagation path is calculated based on the multipath superposition effect.

[0083] The signal strength and the channel fading characteristics are comprehensively evaluated. When the comprehensive evaluation result meets the communication quality constraints, it is determined that the corresponding communication node pairs have a reachability relationship, and connection weights are assigned to the communication node pairs with a reachability relationship.

[0084] Based on the topology discovery mechanism, a network topology structure containing node identifiers, inter-node connection relationships, and connection weights is constructed according to the reachability relationship and the connection weight.

[0085] The location information of communication nodes is acquired through an underground positioning system. Each communication node is equipped with an ultra-wideband positioning tag or inertial measurement unit, which reports three-dimensional coordinate data to the location management server in real time. The location data includes a node identifier, X-axis coordinate, Y-axis coordinate, Z-axis coordinate, and a timestamp field, with coordinate accuracy controlled within 0.5 meters. The spatial distance calculation module receives the location information of any two communication nodes and calculates the straight-line distance between the nodes using the three-dimensional Euclidean distance formula. The calculation result is expressed in meters and retained to one decimal place.

[0086] The geometric constraints of the underground roadway are obtained through a 3D roadway model database, which stores geometric parameters such as the roadway centerline orientation, cross-sectional dimensions, turning angles, and bifurcation locations. The propagation path identification module performs ray tracing analysis based on the spatial location of the communication node pairs and the roadway geometric constraints. When the connection between nodes is entirely within the roadway space and not obstructed by the roadway walls or large equipment, it is identified as a straight propagation path and assigned the path identifier "01". When the straight connection between nodes is blocked by roadway turns, bifurcations, or obstacles, and the signal needs to propagate through roadway wall reflection or around obstacles, it is identified as a diffraction propagation path and assigned the path identifier "02".

[0087] Channel state information is acquired in real time by the radio frequency modules of each communication node, including parameters such as received signal strength indication, channel power spectral density, and multipath delay spread. Signal strength measurement uses logarithmic scalar representation, covering a range of -120 to -30 dB / mW, with a sampling period of 200 milliseconds. The topology discovery mechanism employs an active probing approach, with each node periodically sending probe frames to its neighbors. The receiver measures the signal strength of the probe frames and feeds it back to the transmitter, establishing bidirectional channel quality assessment data.

[0088] The channel fading characteristics of a straight propagation path are calculated based on a logarithmic distance path loss model, considering the combined effects of spatial distance, operating frequency, and tunnel environmental factors. The path loss value equals the path loss at the reference distance plus the path loss exponent multiplied by the logarithm of the distance ratio, then multiplied by 10. The reference distance is set to 1 meter, and the reference path loss value is determined according to the operating frequency band: 40 dB for the 2.4 GHz band and 46 dB for the 5.8 GHz band. The path loss exponent is set to a range of 2.2 to 2.8 for underground straight tunnel environments, with the specific value adjusted according to the tunnel cross-sectional dimensions and wall material characteristics.

[0089] The channel fading characteristics of the diffracted propagation path are calculated using a multipath superposition analysis method to identify all propagation paths of the signal from the transmitter to the receiver. The main propagation paths include the primary reflection path, the secondary reflection path, and the diffracted path. The signal amplitude of each path is calculated independently based on the propagation distance, reflection loss, and diffraction loss. The reflection loss is determined based on the reflection coefficient of the tunnel wall material; the reflection coefficient of concrete walls is taken as 0.25, and that of steel plate surfaces as 0.75. The diffraction loss is calculated using Fresnel diffraction theory, considering the degree of obstruction by obstacles relative to the Fresnel ellipsoid. The multipath signals are vector-superimposed at the receiver, taking into account the amplitude and phase differences of each path signal to obtain the total power of the synthesized signal.

[0090] The comprehensive evaluation process compares and verifies the measured signal strength with the calculated channel fading characteristics to assess the fit between the theoretical model and the actual environment. Communication quality constraints include three key indicators: minimum received power threshold, carrier-to-noise ratio threshold, and link margin threshold. The minimum received power threshold is determined based on the modulation scheme and bit error rate requirements; it is set to -95 dB / mW for QPSK modulation and -92 dB / mW for 16QAM modulation. The carrier-to-noise ratio threshold is set to 20 dB to ensure demodulation performance meets data transmission requirements. The link margin threshold is set to 10 dB to allow for tolerances caused by channel fading variations and interference.

[0091] Reachability is determined through a multi-indicator comprehensive scoring system. When the signal strength exceeds the minimum received power threshold, the carrier-to-noise ratio exceeds a set threshold, and the link margin meets the requirements, the corresponding communication node pair is deemed reachable. Connection weight allocation is based on a weighted calculation of signal quality indicators, with signal strength accounting for 40%, carrier-to-noise ratio for 35%, and link margin for 25%. Weight values ​​range from 1 to 100 integers, and the calculation results are normalized to ensure the rationality of the weight distribution.

[0092] The network topology is represented using a weighted undirected graph data structure. Node objects include attributes such as unique identifier, location coordinates, device type, and operating status. Connection objects include attributes such as starting node identifier, destination node identifier, connection weight, propagation path type, and establishment time. The topology management module maintains a complete network connection matrix and supports topology query, update, and route calculation functions. When channel conditions change, connection weights are dynamically adjusted, triggering a topology reconstruction process.

[0093] In one optional implementation, based on the network characteristic parameters, a link quality prediction model integrating network state awareness is constructed. The link quality model is then used to determine the link quality level between each communication node, forming a predicted link quality assessment result, including:

[0094] Connectivity, node distribution density, and topology stability are extracted from network feature parameters, and channel state information and service load information between communication nodes are obtained within a historical time period. Based on the temporal correlation between connectivity, node distribution density, topology stability, channel state information, and service load information, a link quality prediction model with fused network state awareness is constructed.

[0095] By inputting the current network characteristic parameters, channel state information, and traffic load information into the link quality prediction model, the predicted link quality values ​​between communication nodes in the future time period are obtained.

[0096] Based on the predicted link quality values ​​and in conjunction with a preset link quality grading standard, the link quality levels between communication nodes are divided into multiple quality levels, and a quality weight coefficient is assigned to each quality level. The quality weight coefficient reflects the degree of influence of different quality levels on path selection.

[0097] The predicted link quality value is combined with the quality level to form the predicted link quality assessment result.

[0098] The extracted network feature parameters include three core indicators: connectivity, node density, and topological stability. Connectivity represents the probability of a communication link between any two nodes in the network, and can be measured by calculating the average number of neighbors of a node. For any node i, its connectivity can be obtained by counting the number of nodes directly connected to it, and the overall network connectivity is the average connectivity of all nodes. Node density reflects the number of communicating nodes per unit area, and can be measured by calculating the number of nodes per unit area or the average distance between nodes. Topological stability represents the frequency of network structure changes, and is calculated by observing the frequency of link establishment and disconnection within a specific time window. Higher stability indicates less topological change.

[0099] The system acquires channel status and service load information among communication nodes over a historical time period. Channel status information includes parameters such as signal-to-noise ratio, bit error rate, and packet loss rate, which can be obtained through periodically exchanged control messages between nodes. Service load information includes node data transmission rate, buffer occupancy rate, and processing latency, obtained by monitoring node processing capacity and resource usage. This information is typically stored in time-series format for easy analysis of its temporal variation patterns.

[0100] When constructing a network state-aware link quality prediction model, a Long Short-Term Memory (LSTM) network structure is used to handle the temporal correlations between the aforementioned parameters. The model consists of three parts: an input layer, a hidden layer, and an output layer. The input layer receives multi-dimensional features such as connectivity, node density, topology stability, channel state information, and traffic load information. The hidden layer consists of multiple LSTM units, each containing an input gate, a forget gate, and an output gate, used to learn short-term and long-term dependencies. The output layer outputs the predicted link quality value. The model is trained using historical datasets, and parameter optimization is performed by minimizing the mean squared error between the predicted values ​​and the actual link quality.

[0101] By inputting the current network characteristic parameters, channel state information, and traffic load information into a trained link quality prediction model, the predicted link quality values ​​between communication nodes in the future time period can be obtained. During the prediction process, the input data is first standardized to ensure consistency in the range of feature data across different dimensions; then, the processed data is input into the model in chronological order; finally, the prediction results are obtained through forward propagation of the model. The predicted values ​​can be expressed as the quality parameters of each communication link at a specific future time point (e.g., 10 minutes, 30 minutes, or 1 hour), including expected bandwidth, latency, and reliability.

[0102] Based on predicted link quality values ​​and pre-defined link quality grading standards, the link quality levels between communication nodes are divided into multiple quality levels. Quality grading is typically set to four levels: Excellent, Good, Average, and Poor. Specific grading standards can be set according to application scenario requirements. For example, in scenarios with low latency requirements, links with end-to-end latency less than 10 milliseconds can be classified as Excellent, 10-50 milliseconds as Good, 50-100 milliseconds as Average, and over 100 milliseconds as Poor. For different quality levels, corresponding quality weight coefficients are assigned; for example, Excellent has a weight of 1.0, Good has a weight of 0.8, Average has a weight of 0.5, and Poor has a weight of 0.2. The quality weight coefficients reflect the degree of influence of different quality levels on path selection; a higher weight indicates a higher priority for the link in routing decisions.

[0103] The predicted link quality value is combined with the quality level to form the predicted link quality assessment result. The assessment result can be represented as a data structure containing multiple attributes, including link identifier, starting node, ending node, predicted quality value, quality level, and quality weight coefficient. This structured representation facilitates network management systems in understanding and utilizing link quality information for operations such as route optimization, load balancing, and resource scheduling.

[0104] In practical applications, taking a wireless mesh network as an example, node A forms multiple communication links with surrounding nodes B, C, and D. Using the model described above to predict the quality of each link over the next 30 minutes, the predicted quality values ​​are: AB link 0.85 (excellent, weight 1.0), AC link 0.62 (good, weight 0.8), and AD link 0.38 (average, weight 0.5). Based on these evaluation results, AB links can be prioritized for data transmission. When network congestion occurs or link status changes, AC or AD links can be considered, thus achieving efficient and reliable network communication.

[0105] In one optional implementation, the communication network is dynamically partitioned by combining the predicted link quality assessment results with the service load information, the transmission nodes in each partition are determined, and a node priority ranking table is established. The optimal transmission path is selected based on the node priority ranking table, including:

[0106] Based on the predicted link quality values ​​in the predicted link quality assessment results, communication node connections in the network topology with predicted link quality values ​​lower than a preset quality threshold are identified. In-depth analysis of communication node connections is performed in conjunction with the link awareness mechanism, and the analysis results are used as partition boundary markers to dynamically partition the communication network.

[0107] Based on the link-aware mechanism, the status of each network partition is continuously monitored, the service load information and location information of each communication node are extracted, and the communication nodes with loads below the load threshold are identified as candidate transmission nodes for the network partition through the resource scheduling algorithm.

[0108] For candidate transmission nodes in each network partition, based on the resource scheduling algorithm and the predicted link quality assessment results and service load information, calculate the comprehensive path quality score and remaining transmission capacity of the node, and establish a node priority ranking table based on the score results;

[0109] Based on the node priority ranking table, the transmission node with the highest ranking and that meets the monitoring requirements of the link awareness mechanism is selected as the relay node to construct the optimal transmission path from the source node to the data aggregation node.

[0110] like Figure 2 As shown, the method includes:

[0111] The predicted link quality assessment results are obtained through a quality prediction module, which maintains a database of predicted quality values ​​for each communication node connection. The predicted quality values ​​are represented numerically from 0 to 100, with higher values ​​indicating better link quality. Preset quality thresholds are configured based on the network application scenario: 75 for real-time data transmission, 60 for general data transmission, and 45 for low-priority data transmission. A quality threshold comparison module compares the predicted quality values ​​for each communication node connection in the network topology. When a connection's predicted quality value is lower than the corresponding preset quality threshold, the connection is marked as a low-quality connection and its identifier is recorded.

[0112] The link awareness mechanism is implemented through a distributed monitoring agent. Each communication node deploys a link status monitoring module to collect local link status parameters such as signal strength, bit error rate, and latency jitter in real time. The sampling period for monitoring parameters is set to 500 milliseconds, and the continuous monitoring time window is 30 seconds. The deep analysis module receives the status data provided by the link awareness mechanism and performs multi-dimensional analysis on low-quality connections, including signal attenuation trends, interference source identification, and propagation path stability assessment. The analysis results are quantified as a connection reliability score, ranging from 1 to 10. Connections with scores below 3 are marked as partition boundaries.

[0113] The dynamic partitioning module divides the communication network into regions based on partition boundary identifiers, using the connected component decomposition algorithm from graph theory. After removing the connections marked as partition boundaries, the original network topology is decomposed into several connected subgraphs, each constituting a network partition. Partition identifiers are encoded using incrementing integers, and partition information is stored in a partition management table containing fields such as partition identifier, node list, and boundary connection list. Partition reconstruction is triggered when the number of newly added low-quality connections exceeds 20% of the total number of network connections or when the number of nodes within a partition is less than 3.

[0114] The link awareness mechanism continuously monitors data using a combination of heartbeat detection and status reporting. Each network partition has a monitoring node responsible for collecting status information from all communication nodes within that partition. Status monitoring parameters include system resource metrics such as CPU utilization, memory usage, network interface traffic, and cache queue length. The service load information extraction module calculates service-related metrics for each communication node, such as packet processing rate, number of concurrent connections, and bandwidth utilization. Location information is obtained through the node positioning module, including spatial attributes such as the node's three-dimensional coordinates, movement speed, and signal coverage area.

[0115] The resource scheduling algorithm employs a load balancing strategy to identify candidate transmission nodes. The algorithm's input parameters include node service load information, hardware performance parameters, and network connection status. The load threshold is dynamically adjusted based on the node's hardware configuration: 80% for high-performance nodes, 70% for medium-performance nodes, and 60% for low-performance nodes. Candidate transmission node selection criteria include current load below the load threshold, remaining memory capacity greater than 100MB, and available network interface bandwidth exceeding 10Mbps. The selection results are stored in a candidate node table, recording information such as node identifier, load level, and available resources.

[0116] The path quality comprehensive score calculation module combines the predicted link quality assessment results with service load information for weighted calculation. The quality score components include: average link quality from the node to its neighbors (40%), remaining node processing capacity (30%), node location advantage (20%), and historical node stability (10%). The link quality component is obtained by averaging the predicted quality values ​​of the node and all its neighbors, with a value ranging from 0 to 100. The processing capacity component is calculated based on the reciprocal of the node's CPU utilization and memory utilization; lower load results in a higher score. The location advantage component considers the node's centrality and connectivity within the partition; nodes in central locations with a large number of connections receive higher scores.

[0117] Remaining transmission capacity is calculated based on the node's network interface parameters and current traffic statistics. The transmission capacity assessment module obtains parameters such as the node's network interface theoretical bandwidth, current traffic, and queue occupancy rate to calculate the available transmission capacity. Available capacity equals the theoretical bandwidth minus the currently occupied bandwidth, multiplied by the queue availability rate. The result is expressed in Mbps with one decimal place. A capacity reservation mechanism reserves 20% of bandwidth resources for critical business traffic to ensure quality of service for important data transmissions.

[0118] The node priority ranking table is built based on the comprehensive score results, and uses a descending order to place the nodes with the highest scores at the top. The ranking table data structure includes fields such as node identifier, comprehensive score, remaining transmission capacity, and update timestamp. The ranking update cycle is set to 10 seconds, triggering an immediate update when a node's status changes significantly. The ranking table maintenance module supports dynamic insertion, deletion, and modification operations to ensure the real-time nature and accuracy of the ranking results.

[0119] The transmission node selection module selects the optimal transmission node as a relay node based on a node priority ranking table. The selection strategy prioritizes the highest-ranked node while verifying whether this node meets the monitoring requirements of the link-aware mechanism. These monitoring requirements include a node response time of less than 100 milliseconds, a link stability score greater than 7, and predicted connection quality values ​​with adjacent nodes all exceeding preset quality thresholds. If the highest-ranked node does not meet the monitoring requirements, the module checks the next node in the ranking table sequentially until a suitable relay node is found.

[0120] The optimal transmission path is constructed using Dijkstra's algorithm, which searches for paths based on node priority ranking. The path search module starts from the source node and expands the search range step-by-step through relay nodes until it reaches the data aggregation node. Path quality evaluation considers multiple factors such as path length, relay node quality, and link stability, selecting the transmission path with the best overall performance. Path information is stored in a path table, including attributes such as path identifier, node sequence, path quality score, and establishment time.

[0121] In one optional implementation, based on the resource scheduling algorithm combined with the predicted link quality assessment results and service load information, the comprehensive path quality score and remaining transmission capacity of the nodes are calculated, and a node priority ranking table is established based on the score results, including:

[0122] Extract the predicted link quality values ​​from each communication node to the data aggregation node from the predicted link quality assessment results. Based on the resource scheduling algorithm, the predicted link quality values ​​and the path transmission hop count are weighted and fused to calculate the comprehensive path quality score from each communication node to the data aggregation node.

[0123] Obtain the service load information of each communication node, wherein the service load information includes the current cache queue length of the node and the maximum cache capacity of the node;

[0124] The resource scheduling algorithm determines the remaining transmission capacity of each communication node based on the ratio of the current cache queue length of the node to the maximum cache capacity of the node, and multiplies the remaining transmission capacity as a load balancing factor with the comprehensive path quality score to obtain a comprehensive node score that takes into account both path quality and node load.

[0125] Based on the numerical value of the node's comprehensive score, the communication nodes are arranged in descending order to form a node priority ranking table that reflects the node's transmission priority level.

[0126] The predicted link quality assessment result extraction module retrieves the predicted link quality values ​​from each communication node to the data aggregation node from the quality prediction database. The predicted quality values ​​are represented by a range of 0 to 100, with higher values ​​indicating better link quality. The extraction process queries the corresponding quality prediction record using the node identifier to obtain key information such as the predicted value, confidence level, and update timestamp. The data extraction interface supports batch query operations, handling up to 500 nodes' quality prediction value retrieval requests in a single query. The validity period of the quality prediction values ​​is set to 300 seconds; data exceeding this period is automatically marked as expired and triggers a re-prediction process.

[0127] The resource scheduling algorithm employs a weighted fusion mechanism to comprehensively evaluate link quality predictions and path hop counts. The path hop count, calculated using the shortest path algorithm, represents the number of relay nodes required for a communication node to reach the data aggregation node. The hop count calculation module maintains the network topology connection matrix and uses a breadth-first search algorithm to determine the shortest path and count the hops. In the weighted fusion calculation, the weight coefficient for link quality predictions is set to 0.7, and the weight coefficient for path hop counts is set to 0.3. Quality predictions are directly used in the calculation, while hop counts undergo a reciprocal transformation and normalization to ensure that lower hop counts correspond to higher scores.

[0128] The overall path quality score is calculated using a weighted summation method, with the result ranging from 0 to 100. The specific calculation process involves multiplying the predicted link quality value by 0.7 and adding the reciprocal of the normalized hop count by 0.3. The reciprocal of the normalized hop count is calculated by subtracting the current hop count from the maximum hop count and then dividing by the maximum hop count, ensuring that nodes with fewer hops receive higher path quality scores. The calculation result is retained to one decimal place, with the numerical precision error controlled within 0.1. The overall path quality score is stored in a score data table, containing fields such as node identifier, score value, calculation time, and validity flag.

[0129] The business load information acquisition module collects load data from each communication node through the node status monitoring interface. This monitoring interface uses a RESTful API design, supporting GET queries for node load status and returning load information in JSON format. The current cache queue length of a node is obtained by counting the number of pending data packets in the node's memory buffer, with the unit being the number of data packets, ranging from 0 to the node's maximum queue capacity. The maximum cache capacity of a node is determined based on its hardware configuration: 10,000 data packets for high-performance nodes, 5,000 data packets for medium-performance nodes, and 2,000 data packets for low-performance nodes.

[0130] The cache queue length monitoring employs a real-time sampling mechanism with a sampling period of 200 milliseconds. Monitoring data is smoothed using a sliding window with a window length of 10 sampling points, and the average queue length within the window is calculated as the current load metric. Business load information includes fields such as node identifier, current queue length, maximum cache capacity, queue utilization, and sampling timestamp. The data synchronization period for load information is set to 1 second to ensure the real-time nature and accuracy of the load status. An abnormal load detection mechanism triggers an alarm when the queue length exceeds 90% of the maximum capacity, and rejects new data packets when the queue length exceeds the maximum capacity.

[0131] The resource scheduling algorithm calculates the remaining transmission capacity based on the ratio of the node's current cache queue length to its maximum cache capacity. This ratio is calculated by dividing the current cache queue length by the node's maximum cache capacity, resulting in a small value between 0 and 1; a smaller value indicates a lighter node load. The remaining transmission capacity is calculated using a linear inverse proportional relationship: 1 minus the ratio of queue length to maximum capacity, then multiplied by the node's theoretical maximum transmission capacity. The theoretical maximum transmission capacity is determined by the node's network interface bandwidth: 1000Mbps for a Gigabit Ethernet interface and 100Mbps for a 100Mbps Ethernet interface.

[0132] The remaining transmission capacity is calculated in Mbps with integer precision. The capacity calculation module supports dynamic adjustment and real-time updates, automatically recalculating the remaining capacity when the node load status changes. The load balancing factor is calculated as the ratio of the remaining transmission capacity to the theoretical maximum transmission capacity, ranging from 0 to 1. A higher ratio indicates more available resources for the node. The load balancing factor is used as a weighting parameter in subsequent comprehensive scoring calculations to ensure balanced resource utilization.

[0133] The node's overall score is calculated by multiplying the load balancing factor and the path quality score. This multiplication ensures a comprehensive consideration of both path quality and node load, with nodes possessing high path quality and low load receiving the highest overall score. The overall score ranges from 0 to 100, with calculation precision rounded to one decimal place. The score calculation module employs parallel processing to improve computational efficiency, supporting simultaneous score calculations for up to 1000 nodes. The calculation results are updated in real-time and stored in the overall score database, containing fields such as node identifier, overall score, path quality score, load balancing factor, and update time.

[0134] The node priority sorting table is constructed based on the numerical values ​​of the node's comprehensive score, arranged in descending order. The sorting algorithm uses quicksort, with a time complexity of O(nlogn), suitable for large-scale node sorting scenarios. The sorting table data structure uses a dynamic array for storage, supporting efficient insertion, deletion, and search operations. The sorting table includes fields such as sort order number, node identifier, comprehensive score, path quality score, remaining transmission capacity, and load balancing factor. The sorting update strategy uses incremental updates; when a node's score changes, only the affected sorting position is adjusted locally, avoiding the performance overhead of a full reordering.

[0135] The sorting table maintenance mechanism supports dynamic addition and deletion of nodes. When a new node is added, it is inserted into the corresponding sorting position based on its comprehensive score. When a node is deleted, the corresponding record is removed from the sorting table, and the sorting order of subsequent nodes is adjusted. The storage capacity of the sorting table is set to 10,000 node records. When the capacity is exceeded, a lowest score eviction strategy is used to maintain the table capacity. The sorting table access interface provides multiple query methods, such as querying by sequence number range, filtering by score threshold, and locating by node identifier, to meet the needs of different application scenarios.

[0136] In one optional implementation, recalculating the target path based on the updated network characteristic parameters and establishing a communication link to perform data transmission includes:

[0137] The updated network feature parameters are obtained, and the network status change trend of each communication node in the network feature parameters is analyzed. Based on the network status change trend, the weight coefficient of each communication node in the target path selection is dynamically adjusted to construct a path optimization strategy that reflects the dynamic adaptability of the network. Based on the path optimization strategy, the target path is determined, and a communication link is established to execute data transmission.

[0138] Acquire updated network characteristic parameters, which typically include key indicators such as link latency, bandwidth utilization, node load, packet loss rate, and signal strength. In the network monitoring system, set the data acquisition cycle to T (usually several seconds to minutes) to ensure timely capture of network changes. Once a new round of network characteristic parameters is collected, store it in the network status database, while also retaining historical data records for subsequent trend analysis.

[0139] Analyzing the network state change trends of each communication node among the network characteristic parameters involves using time series analysis to calculate the rate and direction of change of each network parameter for each communication node. Specifically, for a network parameter P of node i, the slope of its change over multiple consecutive time windows is calculated, forming a trend vector V_i. For example, if the bandwidth utilization of a node has been continuously increasing over the past five sampling periods, it is determined that the node's bandwidth resources are approaching saturation; if the packet loss rate shows a fluctuating upward trend, it indicates that the stability of the node is decreasing.

[0140] Based on trend analysis, the weight coefficients of each communication node in target path selection are dynamically adjusted. The weight adjustment employs an adaptive weight calculation method, comprehensively evaluating the reliability, stability, and performance of each node according to network state change trends. If a node's network state shows a deteriorating trend (e.g., increased latency, higher packet loss rate), its weight coefficient is reduced; conversely, if a node's state shows an improving trend, its weight is increased. Specifically, a trend coefficient correction method can be used, multiplying the original weight W_i of node i by a trend correction factor to obtain the adjusted weight W_i'. The trend correction factor is calculated based on the magnitude and direction of changes in each parameter in the trend vector V_i.

[0141] A path selection strategy reflecting the network's dynamic adaptability is constructed. This strategy comprehensively considers adjusted node weights and link states to build a weighted graph of the network topology. In this weighted graph, node weights represent the reliability and processing capacity of communication nodes, while edge weights represent the transmission performance of links. Furthermore, the path selection strategy incorporates a predictive evaluation mechanism, predicting the state changes of each node and link in the near future based on historical trends, and prioritizing path combinations that are not only currently in good condition but also have stable or positive future trends.

[0142] In one embodiment, a data transmission link needs to be established between nodes A and J in a network consisting of 10 communication nodes. Initial monitoring revealed that the load on node C was rapidly increasing, with bandwidth utilization rising from 60% to 85% over the past three sampling periods, while node D, although currently heavily loaded (70%), showed a steady downward trend. After applying a dynamic weight adjustment mechanism, the weight coefficient of node C decreased from 0.8 to 0.6, while the weight of node D increased from 0.7 to 0.8. The final calculated target path avoided node C, selecting a path passing through node D, namely A→B→D→F→J, rather than the seemingly shorter A→C→E→J path.

[0143] Based on the constructed path optimization strategy, a shortest path algorithm is used to determine the final target path. This algorithm not only considers traditional path length metrics but also incorporates adjusted node weights and predictive evaluation results. By iteratively comparing the comprehensive scores of multiple candidate paths, the path with the best overall performance is selected as the target path. The comprehensive score calculation formula comprehensively considers multiple metrics such as path latency, reliability, load balancing, and future trend assessment.

[0144] Based on the determined target path, a communication link is established and data transmission is executed. This stage begins by sending path confirmation information to each node on the target path to verify their current availability. Once confirmed, an end-to-end communication link is established, necessary resources (such as buffers and processing time slices) are allocated, and data transmission begins. During data transmission, the link status is continuously monitored. If a significant deviation between actual transmission performance and expectations is detected, an emergency path recalculation mechanism is triggered to ensure the reliability and efficiency of data transmission.

[0145] After data transmission is completed, the actual network performance metrics obtained during the transmission process are compared and analyzed with the predicted metrics. The parameters of the prediction model are then adjusted to improve the accuracy of future predictions. This closed-loop optimization mechanism enables the path selection strategy to continuously improve itself and adapt to dynamic changes in the network environment.

[0146] The above technical solution enables adaptive path selection based on dynamic changes in network characteristic parameters, effectively improving the reliability and efficiency of data transmission, and is particularly suitable for complex communication environments where network conditions change frequently.

[0147] A second aspect of the present invention provides a downhole communication link optimization system based on a self-organizing network, comprising:

[0148] The first unit is used to acquire the location information, channel status information, and service load information of multiple communication nodes distributed underground.

[0149] The second unit is used to construct a network topology structure reflecting the connectivity between nodes based on the location information and the channel state information through a topology discovery mechanism, and to analyze the connectivity, node distribution density and topology stability of the network topology structure to determine network characteristic parameters.

[0150] The third unit is used to construct a link quality prediction model that integrates network state awareness based on the network feature parameters, and to use the link quality model to determine the link quality level between each communication node, thereby forming a predicted link quality assessment result.

[0151] The fourth unit is used to combine the predicted link quality assessment results with the service load information to dynamically partition the communication network, determine the transmission nodes in each partition and establish a node priority ranking table, select the optimal transmission path according to the node priority ranking table, periodically detect the link status of the optimal transmission path, and when a reduction in link transmission quality is detected and an adjustment threshold is triggered, activate the node self-organizing reconstruction process and re-execute the topology discovery mechanism to update the network characteristic parameters.

[0152] The fifth unit is used to recalculate the target path based on the updated network feature parameters and establish a communication link to perform data transmission.

[0153] A third aspect of the present invention provides an electronic device, comprising:

[0154] processor;

[0155] Memory used to store processor-executable instructions;

[0156] The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0157] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0158] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0159] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing downhole communication links based on self-organizing networks, characterized in that, include: Acquire location information, channel status information, and service load information of multiple communication nodes distributed underground; Based on the location information and the channel state information, a network topology structure reflecting the connectivity between nodes is constructed through a topology discovery mechanism. The connectivity, node distribution density and topology stability of the network topology structure are analyzed to determine network characteristic parameters. Based on the network characteristic parameters, a link quality prediction model integrating network state awareness is constructed. The link quality prediction model is used to determine the link quality level between each communication node, and a predicted link quality assessment result is formed. Combining the predicted link quality assessment results with the service load information, the communication network is dynamically partitioned, the transmission nodes in each partition are determined, and a node priority ranking table is established. The optimal transmission path is selected according to the node priority ranking table. The link status of the optimal transmission path is periodically probed. When a reduction in link transmission quality is detected and an adjustment threshold is triggered, the node self-organizing reconstruction process is activated and the topology discovery mechanism is re-executed to update the network characteristic parameters. The target path is recalculated based on the updated network feature parameters, and a communication link is established to perform data transmission. The step of combining the predicted link quality assessment results with the service load information to dynamically partition the communication network, determine the transmission nodes in each partition and establish a node priority ranking table, and select the optimal transmission path according to the node priority ranking table includes: Based on the predicted link quality values ​​in the predicted link quality assessment results, communication node connections in the network topology with predicted link quality values ​​lower than a preset quality threshold are identified. In-depth analysis of communication node connections is performed in conjunction with the link awareness mechanism, and the analysis results are used as partition boundary markers to dynamically partition the communication network. Based on the link-aware mechanism, the status of each network partition is continuously monitored, the service load information and location information of each communication node are extracted, and the communication nodes with loads below the load threshold are identified as candidate transmission nodes for the network partition through the resource scheduling algorithm. For candidate transmission nodes in each network partition, based on the resource scheduling algorithm and the predicted link quality assessment results and service load information, calculate the comprehensive path quality score and remaining transmission capacity of the node, and establish a node priority ranking table based on the score results; Based on the node priority ranking table, the transmission node with the highest ranking and that meets the monitoring requirements of the link awareness mechanism is selected as the relay node to construct the optimal transmission path from the source node to the data aggregation node. The step of calculating the comprehensive path quality score and remaining transmission capacity of a node based on the resource scheduling algorithm, combined with the predicted link quality assessment results and service load information, and establishing a node priority ranking table based on the score results includes: Extract the predicted link quality values ​​from each communication node to the data aggregation node from the predicted link quality assessment results. Based on the resource scheduling algorithm, the predicted link quality values ​​and the path transmission hop count are weighted and fused to calculate the comprehensive path quality score from each communication node to the data aggregation node. Obtain the service load information of each communication node, wherein the service load information includes the current cache queue length of the node and the maximum cache capacity of the node; The resource scheduling algorithm determines the remaining transmission capacity of each communication node based on the ratio of the current cache queue length of the node to the maximum cache capacity of the node, and multiplies the remaining transmission capacity as a load balancing factor with the comprehensive path quality score to obtain a comprehensive node score that takes into account both path quality and node load. Based on the numerical value of the node's comprehensive score, the communication nodes are arranged in descending order to form a node priority ranking table that reflects the node's transmission priority level.

2. The method according to claim 1, characterized in that, Based on the location information and the channel state information, a network topology reflecting the connectivity relationships between nodes is constructed through a topology discovery mechanism. The network topology's connectivity, node density, and topology stability are analyzed to determine network characteristic parameters, including: Based on location information, the spatial distance between each communication node is calculated. Combined with the signal strength and channel fading characteristics in the channel state information, the reachability relationship between each communication node is determined through the topology discovery mechanism, and a network topology structure containing node identifiers and inter-node connection relationships is constructed. During the execution of the topology discovery mechanism, neighbor discovery response information of each communication node is collected and the number of neighbor nodes is counted. The connectivity degree, which reflects the overall connectivity of the network, is calculated by combining the number of connected subgraphs in the network. Based on the location information, the underground space is divided into multiple regional units. The distribution of communication nodes in each regional unit is statistically analyzed. Based on the variance analysis of the regional node distribution, the node distribution density reflecting the uniformity of spatial distribution is determined. By periodically executing the topology discovery mechanism, the changes in the connection relationships between nodes within a preset time window are recorded, and the topology stability, which reflects the degree of dynamic change in the network topology, is calculated based on the frequency of changes in the connection relationships. The connectivity, node distribution density, and topological stability are subjected to hierarchical weighted mapping, and dynamic adjustment coefficients are set according to the importance of each parameter to form network feature parameters.

3. The method according to claim 2, characterized in that, Based on location information, the spatial distance between communication nodes is calculated. Combined with signal strength and channel fading characteristics from channel state information, a topology discovery mechanism is used to determine the reachability relationships between communication nodes. This constructs a network topology structure that includes node identifiers and inter-node connectivity relationships, including: The spatial distance between each communication node is calculated based on the location information. Based on the spatial distance and the geometric constraint relationship between the underground roadway, communication node pairs with straight propagation paths and communication node pairs with diffraction propagation paths are identified to obtain propagation path type identifiers. The topology discovery mechanism is used to calculate the channel fading characteristics of the straight propagation path and the diffraction propagation path based on the signal strength in the channel state information and the propagation path type identifier. The straight propagation path is calculated based on the spatial distance attenuation characteristics, and the diffraction propagation path is calculated based on the multipath superposition effect. The signal strength and the channel fading characteristics are comprehensively evaluated. When the comprehensive evaluation result meets the communication quality constraints, it is determined that the corresponding communication node pairs have a reachability relationship, and connection weights are assigned to the communication node pairs with a reachability relationship. Based on the topology discovery mechanism, a network topology structure containing node identifiers, inter-node connection relationships, and connection weights is constructed according to the reachability relationship and the connection weight.

4. The method according to claim 1, characterized in that, Based on the network characteristic parameters, a link quality prediction model integrating network state awareness is constructed. This model is then used to determine the link quality level between each communication node, resulting in a predicted link quality assessment result, including: Connectivity, node distribution density, and topology stability are extracted from network feature parameters, and channel state information and service load information between communication nodes are obtained within a historical time period. Based on the temporal correlation between connectivity, node distribution density, topology stability, channel state information, and service load information, a link quality prediction model with fused network state awareness is constructed. By inputting the current network characteristic parameters, channel state information, and traffic load information into the link quality prediction model, the predicted link quality values ​​between communication nodes in the future time period are obtained. Based on the predicted link quality values ​​and in conjunction with a preset link quality grading standard, the link quality levels between communication nodes are divided into multiple quality levels, and a quality weight coefficient is assigned to each quality level. The quality weight coefficient reflects the degree of influence of different quality levels on path selection. The predicted link quality value is combined with the quality level to form the predicted link quality assessment result.

5. The method according to claim 1, characterized in that, The process of recalculating the target path based on the updated network feature parameters and establishing a communication link to perform data transmission includes: The updated network feature parameters are obtained, and the network status change trend of each communication node in the network feature parameters is analyzed. Based on the network status change trend, the weight coefficient of each communication node in the target path selection is dynamically adjusted to construct a path optimization strategy that reflects the dynamic adaptability of the network. Based on the path optimization strategy, the target path is determined, and a communication link is established to execute data transmission.

6. A downhole communication link optimization system based on self-organizing networks, used to implement the method of any one of claims 1-5, characterized in that, include: The first unit is used to acquire the location information, channel status information, and service load information of multiple communication nodes distributed underground. The second unit is used to construct a network topology structure reflecting the connectivity between nodes based on the location information and the channel state information through a topology discovery mechanism, and to analyze the connectivity, node distribution density and topology stability of the network topology structure to determine network characteristic parameters. The third unit is used to construct a link quality prediction model that integrates network state awareness based on the network feature parameters, and to use the link quality prediction model to determine the link quality level between each communication node, thereby forming a predicted link quality assessment result. The fourth unit is used to combine the predicted link quality assessment results with the service load information to dynamically partition the communication network, determine the transmission nodes in each partition and establish a node priority ranking table, select the optimal transmission path according to the node priority ranking table, periodically detect the link status of the optimal transmission path, and when a reduction in link transmission quality is detected and an adjustment threshold is triggered, activate the node self-organizing reconstruction process and re-execute the topology discovery mechanism to update the network characteristic parameters. The fifth unit is used to recalculate the target path based on the updated network feature parameters and establish a communication link to perform data transmission; The step of combining the predicted link quality assessment results with the service load information to dynamically partition the communication network, determine the transmission nodes in each partition and establish a node priority ranking table, and select the optimal transmission path according to the node priority ranking table includes: Based on the predicted link quality values ​​in the predicted link quality assessment results, communication node connections in the network topology with predicted link quality values ​​lower than a preset quality threshold are identified. In-depth analysis of communication node connections is performed in conjunction with the link awareness mechanism, and the analysis results are used as partition boundary markers to dynamically partition the communication network. Based on the link-aware mechanism, the status of each network partition is continuously monitored, the service load information and location information of each communication node are extracted, and the communication nodes with loads below the load threshold are identified as candidate transmission nodes for the network partition through the resource scheduling algorithm. For candidate transmission nodes in each network partition, based on the resource scheduling algorithm and the predicted link quality assessment results and service load information, calculate the comprehensive path quality score and remaining transmission capacity of the node, and establish a node priority ranking table based on the score results; Based on the node priority ranking table, the transmission node with the highest ranking and that meets the monitoring requirements of the link awareness mechanism is selected as the relay node to construct the optimal transmission path from the source node to the data aggregation node. The step of calculating the comprehensive path quality score and remaining transmission capacity of a node based on the resource scheduling algorithm, combined with the predicted link quality assessment results and service load information, and establishing a node priority ranking table based on the score results includes: Extract the predicted link quality values ​​from each communication node to the data aggregation node from the predicted link quality assessment results. Based on the resource scheduling algorithm, the predicted link quality values ​​and the path transmission hop count are weighted and fused to calculate the comprehensive path quality score from each communication node to the data aggregation node. Obtain the service load information of each communication node, wherein the service load information includes the current cache queue length of the node and the maximum cache capacity of the node; The resource scheduling algorithm determines the remaining transmission capacity of each communication node based on the ratio of the current cache queue length of the node to the maximum cache capacity of the node, and multiplies the remaining transmission capacity as a load balancing factor with the comprehensive path quality score to obtain a comprehensive node score that takes into account both path quality and node load. Based on the numerical value of the node's comprehensive score, the communication nodes are arranged in descending order to form a node priority ranking table that reflects the node's transmission priority level.

7. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 5.

8. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 5.

Images