Dynamic topology ad hoc network method based on environment perception and adaptive optimization

Abstract

The invention discloses a dynamic topology ad hoc network method based on environmental perception and adaptive optimization, and relates to the field of wireless sensor networks. Parameters (such as temperature, obstacle distribution and node energy consumption) of nodes are collected in real time; a multi-dimensional quantization parameter matrix including an obstacle coverage matrix, an energy consumption distribution matrix and a signal attenuation coefficient matrix is constructed, and path loss, communication delay and node energy consumption are fused and calculated to generate a node weight value by adopting an adaptive weight calculation mechanism, so that the priority of a communication link is dynamically adjusted. Compared with the prior art, the method has the advantages that the self-healing capability and the real-time performance of the network in a complex environment are remarkably improved, the global calculation overhead is reduced, the real-time performance and the expansibility of the system are improved, the node layout and the dynamic optimization of a communication path are realized, and the network load and the stability are balanced.

Description

Technical Field

[0001] The present invention relates to the field of wireless sensor networks, and in particular to a dynamic topology self-organizing network method based on environment perception and adaptive optimization. Background Art

[0002] With the development of Internet of Things (IoT) technology, wireless sensor networks (WSNs) have been widely used in fields such as environmental monitoring and security surveillance. However, in complex environments, nodes are susceptible to obstruction, signal interference, or uneven energy consumption, resulting in reduced network coverage and increased communication latency.

[0003] Traditional topology optimization algorithms use static configuration or global adjustment, lack adaptability, and are unable to meet the needs of real-time changing environments.

[0004] Patent CN105992301A proposes an ad hoc network system based on static formation clustering. By establishing a virtual routing table between cluster heads, it combines active intra-cluster routing with inter-cluster virtual routing, ensuring real-time performance while reducing control signaling overhead. This system leverages the stable topology of static formations, forwarding cluster head messages through gateway nodes to establish virtual connections between cluster heads, thereby optimizing cross-cluster routing, reducing energy consumption, and improving response efficiency. Compared to traditional static clustering methods, its innovation lies in the dynamic maintenance of virtual routing tables, which enhances adaptability to network changes. However, this method relies on a static formation structure and its ability to respond to real-time environmental changes is still limited.

[0005] Furthermore, patent CN115413041B proposes a centralized wireless ad hoc network resource allocation method, which uses a central node to uniformly schedule resources and dynamically adjusts the allocation strategy based on service priority and data volume. When the total resource demand exceeds the available resources, this method uses a weighted calculation of link priority (formula: , where For nodes m link i The weighted value of the last cycle is , the amount of data to be sent is , I is the number of business priorities, 、 、 The three weight coefficients are determined by the specific system), giving priority to the transmission resources of high-priority services, and introducing the parameter of the amount of data sent and Reduce the "ping-pong" effect of resource allocation. However, this solution relies on centralized computing at a central node, which has problems such as high computational overhead and insufficient scalability, and does not fully consider the impact of dynamic network changes on resource allocation.

[0006] With the rapid development of wireless sensor networks, there is an urgent need for a method that can dynamically perceive the environment and adaptively adjust the topology to improve network performance in complex scenarios. Summary of the Invention

[0007] In response to the above-mentioned deficiencies in the prior art, the present invention provides a dynamic topology self-organizing networking method based on environmental perception and adaptive optimization, which solves the problem that the existing networking resource allocation method has insufficient response capability to real-time environmental changes, making it difficult to balance network coverage and stability.

[0008] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is: A method for dynamic topology self-organizing network based on environment perception and adaptive optimization is provided, which includes the following steps: Collect node parameters in real time and construct a multi-dimensional quantitative parameter matrix including obstacle coverage matrix, energy consumption distribution matrix and signal attenuation coefficient matrix; the node parameters include physical environment parameters, network status parameters and dynamic interference parameters; Based on the multi-dimensional quantitative parameter matrix, the node weight value is generated by integrating the path loss, communication delay and node energy consumption parameters; According to the multi-dimensional quantitative parameter matrix and node weight values, the optimal communication link priority of each node is calculated through game theory strategy to ensure balanced allocation of network resources and complete dynamic topology self-organizing network.

[0009] Furthermore, the physical environment parameters of the node include temperature gradient distribution data and obstacle distribution data. The specific method for collecting the physical environment parameters of the node in real time is: The temperature changes in the area where the node is located are monitored with a sampling period of 10 seconds and an interval of 0.5°C to obtain temperature gradient distribution data; Infrared detection is used to identify and record the coordinates of obstacles with a resolution of 10 cm to obtain obstacle distribution data; The network status parameters of the node include the node remaining energy data and the communication link quality data. The specific method for real-time collection of the node network status parameters is as follows: Monitor the battery power of the node in real time with 1% accuracy to obtain the remaining energy data of the node; Evaluate the signal-to-noise ratio (SNR) of a node. If the SNR is greater than or equal to 15dB, the node is determined to have reliable communication link quality and obtain communication link quality data. During the SNR evaluation process, the physical distance between nodes is calculated based on the signal propagation time. The dynamic interference parameters of a node include the channel congestion level and the signal attenuation coefficient. The specific method for real-time collection of the dynamic interference parameters of a node is as follows: The sampling period is 100ms, and the packet collision rate is used as the channel congestion level; The signal attenuation coefficient is dynamically calculated based on the received signal strength of the node, and its expression is: in Represents the signal attenuation coefficient of a single node; is the transmit power of the transmitting node; is the receiving power of the receiving node; is the distance between the transmitting node and the receiving node; the signal attenuation range is 0.5 dB / m ~3.0 dB / m.

[0010] Furthermore, a specific method for constructing a multi-dimensional quantization parameter matrix including an obstacle coverage matrix, an energy consumption distribution matrix, and a signal attenuation coefficient matrix includes the following steps: The physical environment parameters, network status parameters, and dynamic interference parameters of the nodes collected in real time are filtered for noise, and the target area is divided into 1m×1m grid cells. The physical environment parameters, network status parameters, and dynamic interference parameters of the nodes corresponding to each grid cell are obtained. Divide each grid cell into 100 identical sub-cells. Count the number of sub-cells containing obstacles based on the coordinates of the obstacles. Divide the number of sub-cells containing obstacles in a single grid cell by 100 to obtain the obstacle coverage rate in a single grid cell. The first i Rank j The obstacle coverage rate in the column grid cells is used as the first i Rank j The elements of the column are used to obtain the obstacle coverage matrix; The grid cells with node residual energy lower than 20% are recorded as high energy consumption areas, the grid cells with node residual energy between 20% and 60% are recorded as medium energy consumption areas, and the grid cells with node residual energy greater than 60% are recorded as low energy consumption areas. The energy consumption distribution values ​​of the low energy consumption area, medium energy consumption area, and high energy consumption area are set to 0, 0.5, and 1, respectively, to obtain the energy consumption distribution values ​​corresponding to each grid cell, and then the energy consumption distribution matrix is ​​obtained. The expression of the energy consumption distribution value in a single grid cell is: Indicates the i Rank j The energy consumption distribution value corresponding to the grid cell at the column position, that is, the energy consumption distribution matrix i Rank j The value of the element at the column position; The signal attenuation coefficient corresponding to each grid unit is obtained according to the signal attenuation coefficient of a single node, and the signal attenuation coefficient corresponding to each grid unit is used as an element of the signal attenuation coefficient matrix to obtain the signal attenuation coefficient matrix.

[0011] Furthermore, based on the multi-dimensional quantization parameter matrix, by fusing path loss, communication delay and node energy consumption parameters, a specific method for generating node weight values ​​includes the following steps: Calculate the obstacle attenuation factor between the transmitting node and the receiving node based on the obstacle coverage matrix The obstacle attenuation factor ranges from 1.0 to 5.0. For every 10% increase in obstacle coverage, the obstacle attenuation factor increases by 0.4. The path loss is calculated using the improved Friis free space model, and its expression is: in is the path loss; is the transmit power; is the antenna gain of the transmitting node; is the antenna gain of the receiving node; is the wavelength of the signal carrier; is pi; is the distance between the transmitting node and the receiving node; The communication delay is calculated based on the transmission distance and link utilization, and its expression is: in For communication delay; is the signal propagation speed; is the queue length; is bandwidth; is the link utilization; The node energy consumption is dynamically evaluated based on the remaining energy and energy consumption rate, and its expression is: in is the node energy consumption; is the current remaining energy of the node; is the initial energy of the node; is the amount of energy used; For usage time; Indicates the energy consumption rate; Generate node weight value, the expression is: in is the node weight; 、 and They are path loss weight, communication delay weight and node energy consumption weight respectively.

[0012] Furthermore, when the temperature in the area where the node is located is greater than or equal to 50°C, the The value of is reduced by 0.1, and The value of is increased by 0.15; When the obstacle coverage between the transmitting node and the receiving node is greater than 30%, each 10% increase will The value of increases by 0.2; when When , the queue delay weight of the corresponding node is increased by 50%; When the energy consumption rate of a node is higher than 30% of the average energy consumption rate of the entire network, the node energy consumption weight of the node will be reduced by 20%.

[0013] Furthermore, a specific method for calculating the optimal communication link priority of each node using a game theory strategy based on the multidimensional quantization parameter matrix and the node weight values ​​includes the following steps: Perform preliminary link optimization with the goal of link load balancing through the Nash equilibrium strategy to obtain the initially optimized link; Based on the initially optimized links, the ant colony algorithm is combined with pheromone concentration and node weight value to perform global path optimization with communication efficiency as the goal, update the path selection probability, and obtain the selected path; Based on the selected path, the particle swarm algorithm is used to optimize network coverage and energy consumption balance, adjust the node positions, ensure balanced distribution of network resources, and complete the dynamic topology self-organizing network.

[0014] Furthermore, a preliminary link optimization is performed with the link load balancing as the goal through the Nash equilibrium strategy. The specific method for obtaining the preliminary optimized link is as follows: Dynamically balance node energy consumption weights through inter-node negotiation , so that the standard deviation of energy consumption distribution is less than or equal to 15%; When the standard deviation of the remaining energy of the nodes in the target area exceeds 15%, the energy consumption weight of the nodes corresponding to the high energy consumption area is increased. Reduce by 0.1 and make the node energy consumption weight corresponding to the nodes in the low energy consumption area Increase by 0.1; If the link load difference after preliminary optimization exceeds 20%, the dynamic priority allocation mechanism is triggered to keep the standard deviation of the utilization of each link within 10%.

[0015] Furthermore, the specific method of performing global path optimization with communication efficiency as the goal by combining pheromone concentration and node weight value through ant colony algorithm includes the following steps: Taking the path as an individual of the ant colony, the ant colony algorithm is run with communication efficiency as the goal to obtain the pheromone concentration corresponding to the path; The weight of pheromone concentration is set to 40%, and the weight of node weight is set to 60%. The probability of path selection is calculated as follows: in For path The probability of selection; For path pheromone concentration; For path The node weight value of For path The node weight value of For path pheromone concentration; The path with the highest selection probability is selected as the selected path.

[0016] Furthermore, during the calculation of pheromone concentration, for links with a path loss not exceeding 80dB and a communication delay not exceeding 50ms, the pheromone concentration is increased by an additional 20%; the pheromone volatilization rate is set at 5% / minute.

[0017] Furthermore, the particle swarm algorithm is used to optimize network coverage and energy consumption balance, and the specific method of adjusting the node position by combining the node position and the node weight value is as follows: The node positions are used as particles in the particle swarm algorithm. The particle swarm algorithm is run with the network coverage greater than or equal to 85% and the energy consumption distribution standard deviation less than or equal to 15% as the optimization objectives to obtain the node positions.

[0018] The beneficial effects of this invention are as follows: By collecting node parameters (such as temperature, obstacle distribution, and node energy consumption) in real time, the present invention constructs a multidimensional quantitative parameter matrix consisting of an obstacle coverage matrix, an energy consumption distribution matrix, and a signal attenuation coefficient matrix. Furthermore, an adaptive weight calculation mechanism is used to integrate path loss, communication delay, and node energy consumption to generate node weights, facilitating dynamic adjustment of communication link priorities. Compared to existing technologies, this invention significantly improves the network's self-healing capabilities and real-time performance in complex environments, reduces global computational overhead, enhances system real-time performance and scalability, enables dynamic optimization of node layout and communication paths, and balances network load and stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 Schematic diagram of the process of this method. DETAILED DESCRIPTION

[0020] The specific embodiments of the present invention are described below to facilitate understanding of the present invention by those skilled in the art. However, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the appended claims, these changes are obvious, and all inventions and creations utilizing the concepts of the present invention are protected.

[0021] like Figure 1 As shown, the dynamic topology self-organizing network method based on environment perception and adaptive optimization includes the following steps: S1. Real-time node parameter collection and construction of a multi-dimensional quantitative parameter matrix including an obstacle coverage matrix, an energy consumption distribution matrix, and a signal attenuation coefficient matrix. Node parameters include physical environment parameters, network status parameters, and dynamic interference parameters. S2. Generate node weight values ​​based on the multi-dimensional quantization parameter matrix by integrating path loss, communication delay and node energy consumption parameters; S3. Based on the multi-dimensional quantitative parameter matrix and node weight values, the optimal communication link priority of each node is calculated through game theory strategy to ensure balanced allocation of network resources and complete the dynamic topology self-organizing network.

[0022] In this embodiment, the physical environment parameters of the node include temperature gradient distribution data and obstacle distribution data. The specific method for collecting the physical environment parameters of the node in real time is: The temperature changes in the area where the node is located are monitored with a sampling period of 10 seconds and an interval of 0.5°C to obtain temperature gradient distribution data; Infrared detection is used to identify and record the coordinates of obstacles with a resolution of 10 cm to obtain obstacle distribution data.

[0023] In this embodiment, the network status parameters of the node include node remaining energy data and communication link quality data. The specific method for real-time collection of the network status parameters of the node is: Monitor the battery power of the node in real time with 1% accuracy to obtain the remaining energy data of the node; Evaluate the node's signal-to-noise ratio (SNR). If the SNR is greater than or equal to 15dB, the node is deemed to have reliable communication link quality and obtain communication link quality data. During the SNR evaluation process, the physical distance between nodes is calculated based on signal propagation time with an accuracy of ±0.1m. A link handover is triggered when the SNR is less than 15dB.

[0024] In this embodiment, the dynamic interference parameters of the node include the channel congestion level and the signal attenuation coefficient. The specific method for collecting the dynamic interference parameters of the node in real time is: The sampling period is 100ms, and the packet collision rate is used as the channel congestion level. A link with a collision rate ≥ 15% is marked as highly congested. The signal attenuation coefficient is dynamically calculated based on the received signal strength of the node, and its expression is: in Represents the signal attenuation coefficient of a single node; is the transmit power of the transmitting node; is the receiving power of the receiving node; is the distance between the transmitting node and the receiving node; the signal attenuation range is 0.5 dB / m ~ 3.0 dB / m, that is, when the signal attenuation is less than 0.5 dB / m, the uniform value is 0.5 dB / m; when the signal attenuation is greater than 3.0 dB / m, the uniform value is 3.0 dB / m.

[0025] In this embodiment, the specific method of constructing the multidimensional quantization parameter matrix including the obstacle coverage matrix, the energy consumption distribution matrix, and the signal attenuation coefficient matrix in step S1 includes the following steps: S1-1. Filter out the noise of the node's physical environment parameters, network status parameters, and dynamic interference parameters collected in real time, divide the target area into 1m×1m grid cells, and obtain the physical environment parameters, network status parameters, and dynamic interference parameters of the corresponding node in each grid cell; S1-2. Divide each grid cell into 100 identical sub-cells. Count the number of sub-cells containing obstacles based on the coordinates of the obstacles. Divide the number of sub-cells containing obstacles in a single grid cell by 100 (i.e., the number of sub-cells divided into a single grid cell) to obtain the obstacle coverage rate in the single grid cell. S1-3, the i Rank j The obstacle coverage rate in the column grid cells is used as the first i Rank j The elements of the column are used to obtain the obstacle coverage matrix; S1-4. Grid cells with node residual energy below 20% are recorded as high energy consumption areas, grid cells with node residual energy between 20% and 60% are recorded as medium energy consumption areas, and grid cells with node residual energy greater than 60% are recorded as low energy consumption areas. The energy consumption distribution values ​​of the low energy consumption area, medium energy consumption area, and high energy consumption area are set to 0, 0.5, and 1, respectively, to obtain the energy consumption distribution value corresponding to each grid cell, and then obtain the energy consumption distribution matrix. The expression of the energy consumption distribution value in a single grid cell is: Indicates the i Rank j The energy consumption distribution value corresponding to the grid cell at the column position, that is, the energy consumption distribution matrixi Rank j The value of the element at the column position; S1-5. Obtain the signal attenuation coefficient corresponding to each grid unit according to the signal attenuation coefficient of a single node, and use the signal attenuation coefficient corresponding to each grid unit as an element of a signal attenuation coefficient matrix to obtain a signal attenuation coefficient matrix.

[0026] In step S1-1, noise filtering can be performed by using Kalman filtering to eliminate noise and using fuzzy logic to eliminate outliers to ensure the accuracy and stability of the data.

[0027] For each grid cell, if it contains multiple nodes, the average value of the same type of data of all nodes contained in a single grid cell is used as the corresponding data value. For example, when a single grid cell contains two nodes, the average value of the residual energy of the two nodes is used as the node residual energy of the grid cell.

[0028] In this embodiment, the specific method of generating the node weight value in step S2 based on the multidimensional quantization parameter matrix by fusing the path loss, communication delay and node energy consumption parameters includes the following steps: S2-1. Calculate the obstacle attenuation factor between the transmitting node and the receiving node based on the obstacle coverage matrix The obstacle attenuation factor ranges from 1.0 to 5.0. For every 10% increase in obstacle coverage, the obstacle attenuation factor increases by 0.4. When there are no obstacles between the transmitting node and the receiving node, the obstacle attenuation factor is 1. S2-2. Calculate the path loss using the improved Friis free space model. The expression is: in is the path loss; is the transmit power; is the antenna gain of the transmitting node; is the antenna gain of the receiving node; is the wavelength of the signal carrier; is pi; is the distance between the transmitting node and the receiving node; S2-3. Calculate the communication delay based on the transmission distance and link utilization. The expression is: in For communication delay; is the signal propagation speed; is the queue length; is bandwidth; is the link utilization; S2-4. Dynamically evaluate node energy consumption based on remaining energy and energy consumption rate. The expression is: in is the node energy consumption; is the current remaining energy of the node; is the initial energy of the node; is the amount of energy used; For usage time; Indicates the energy consumption rate. For high energy consumption areas, the energy consumption rate is updated every 5 minutes. S2-5. Generate node weight value, the expression is: in is the node weight; 、 and are path loss weight, communication delay weight and node energy consumption weight respectively, 、 and The initial values ​​of can be set to 0.4, 0.3 and 0.3 respectively.

[0029] In the process of generating node weight values, weights are assigned to path loss, communication delay and node energy consumption respectively, which can optimize the real-time performance of high-congestion areas, avoid high-energy consumption nodes, and extend the network life.

[0030] In this embodiment, when the temperature in the area where the node is located is greater than or equal to 50°C, the temperature will be increased by 5°C for every 5°C increase. The value of is reduced by 0.1 to suppress the dominant influence of path loss on weight and avoid excessive influence of signal attenuation caused by high temperature on link selection. The value of is increased by 0.15 to give priority to low-energy consumption nodes with more remaining energy and extend the network life cycle; When the obstacle coverage between the transmitting node and the receiving node is greater than 30%, each 10% increase will The value of increases by 0.2; when When , the queuing delay weight of the corresponding node is increased by 50% to reflect the additional transmission overhead in high congestion scenarios; When the energy consumption rate of a node is higher than 30% of the average energy consumption rate of the entire network, the node energy consumption weight of the node will be reduced by 20% to avoid high energy consumption hotspots.

[0031] In step S3, the specific method of calculating the optimal communication link priority of each node through a game theory strategy based on the multidimensional quantization parameter matrix and the node weight value includes the following steps: S3-1. Perform preliminary link optimization with link load balancing as the goal using the Nash equilibrium strategy to obtain a preliminarily optimized link. S3-2. Based on the initially optimized links, the ant colony algorithm is used to combine pheromone concentration and node weight values, and global path optimization is performed with communication efficiency as the goal. The path selection probability is updated to obtain the selected path. S3-3. Based on the selected path, the particle swarm algorithm is used to optimize network coverage and energy consumption balance, adjust the node positions, ensure balanced distribution of network resources, and complete the dynamic topology self-organizing network.

[0032] In step S3-1, a preliminary link optimization is performed with the link load balancing as the goal by using the Nash equilibrium strategy. The specific method for obtaining the preliminary optimized link is as follows: Dynamically balance node energy consumption weights through inter-node negotiation (the negotiation frequency in this embodiment is set to once every 10 minutes) , so that the standard deviation of energy consumption distribution is less than or equal to 15%; When the standard deviation of the remaining energy of the nodes in the target area exceeds 15%, the energy consumption weight of the nodes corresponding to the high energy consumption area is increased. Reduce by 0.1 and make the node energy consumption weight corresponding to the nodes in the low energy consumption area Increase by 0.1 to avoid network partition due to energy exhaustion of local nodes; If the link load variance after initial optimization exceeds 20%, a dynamic priority allocation mechanism is triggered to keep the standard deviation of link utilization within 10%. For example, in areas with high obstacle density, nodes will negotiate to lower the priority of high-loss links and redirect traffic to low-latency paths, effectively reducing network congestion.

[0033] In step S3-2, the specific method for performing global path optimization with communication efficiency as the goal by combining the ant colony algorithm with the pheromone concentration and the node weight value includes the following steps: S3-2-1. Take the path as an individual of the ant colony and run the ant colony algorithm with communication efficiency as the goal to obtain the pheromone concentration corresponding to the path; S3-2-2. Set the weight of pheromone concentration to 40% and the weight of node value to 60%, and calculate the probability of path selection. The expression is: in For path The probability of selection; For path pheromone concentration; For path The node weight value of For path The node weight value of For path pheromone concentration; S3-2-3. The path with the highest selection probability is selected as the selected path.

[0034] In specific implementations, global path optimization is performed every five minutes. The ant colony algorithm (ACO) is inspired by the foraging behavior of ants and uses pheromones to achieve group collaboration. Individual ants randomly explore paths, while the colony selects high-probability paths based on pheromone concentration, forming a positive feedback mechanism. The ACO algorithm is used for comparison with existing algorithms. It is not the innovation of this invention and will not be described in detail.

[0035] In this embodiment, during the calculation of pheromone concentration, for links with path loss not exceeding 80dB and communication delay not exceeding 50ms, the pheromone concentration is increased by an additional 20% to guide data flow to a more optimal path; the pheromone volatility rate is set to 5% / minute to prevent outdated path information from affecting real-time decision-making.

[0036] In step S3-3, the particle swarm algorithm is used to optimize network coverage and energy consumption balance. The specific method of adjusting the node position by combining the node position and the node weight value is as follows: The node positions are used as particles in the particle swarm algorithm (PSO), and the particle swarm algorithm is run with the optimization objectives of network coverage greater than or equal to 85% and energy consumption distribution standard deviation less than or equal to 15% to obtain the node positions.

[0037] The particle swarm algorithm (PSO) is inspired by the collective foraging behavior of flocks of birds and schools of fish. Individual particles (particles) collaborate and share information to find the optimal solution, referencing both their own historical best experience (pBest) and tracking the group's global optimal solution (gBest). The PSO algorithm is designed to compare existing algorithms, and the PSO itself is not a novel feature of this invention, so it will not be discussed in detail.

[0038] During implementation, nodes calculate their movement direction based on their current and acquired node positions, with a step size of 0.5m and a maximum movement distance of no more than 10m. Simultaneously, the entire network system monitors coverage blind spots with signal strength below -90dBm in real time, calculates particle velocity vectors, and drives nodes toward the center of the blind spot, ensuring the blind spot is eliminated within three minutes. For example, in energy-intensive areas, the PSO mechanism guides high-energy nodes to migrate to low-coverage areas to balance network load while preventing low-energy nodes from excessively moving and consuming energy.

[0039] In this embodiment, nodes only exchange node parameters and node weights with neighboring nodes within a communication radius of 50m or less. The data exchange cycle is 15 seconds, the single data volume is 1KB or less, and global communication overhead is reduced by 60%. For example, in high-temperature areas, nodes prioritize sharing temperature and energy consumption data, dynamically adjusting local topology weights. Incremental Update: When a new obstacle (distance ≤ 10m), node failure, or link interruption is detected, only local topology reconstruction within a radius of 20m is triggered. A parallel computing framework is used to accelerate the reconstruction process, with a maximum time of 10 seconds or less. A hash check mechanism is used to verify topological consistency and avoid coverage conflicts. For example, after the local reconstruction is completed, the node synchronously updates the obstacle mapping matrix and recalculates the weight values ​​of the affected paths.

[0040] In a specific implementation of the present invention, in a high temperature test scenario (50°C, obstacle density 40%), d = v prop × t , v prop =3×10 8 m / s, transmission power P t =20dBm, antenna gain G t =G r =2dBi, bandwidth B =10Mbps, the queue length threshold is set to 50 packets, the verification coverage rate is improved from 75% to 87%, the average delay is reduced from 120ms to 95ms, the standard deviation of energy consumption balance is reduced from 18% to 11%, and the network lifecycle is extended by 25%.

[0041] In summary, through the collaboration of multiple mechanisms, the present invention can increase the network coverage of the present invention to greater than or equal to 85% in complex scenarios such as high temperature (≥50°C) and high obstacle density (≥30%), and the efficiency of eliminating coverage blind spots is improved by 40%; the average communication delay is stabilized at less than or equal to 100ms, and the difference in high-load link utilization is controlled within ±10%; the topology reconstruction response time is less than or equal to 15 seconds, and the local update resource consumption is reduced by 70%; the standard deviation of node energy consumption balance is less than or equal to 12%, and the network life cycle is extended by 30%, meeting the stringent requirements of high-real-time scenarios such as industrial monitoring and disaster relief.

Claims

1. A dynamic topology self-organizing network method based on environment perception and adaptive optimization, characterized in that: The following steps are involved: Collect node parameters in real time and construct a multi-dimensional quantitative parameter matrix including obstacle coverage matrix, energy consumption distribution matrix and signal attenuation coefficient matrix; The node parameters include physical environment parameters, network status parameters and dynamic interference parameters; Based on the multi-dimensional quantitative parameter matrix, the node weight value is generated by integrating the path loss, communication delay and node energy consumption parameters; According to the multi-dimensional quantitative parameter matrix and node weight values, the optimal communication link priority of each node is calculated through game theory strategy to ensure balanced allocation of network resources and complete dynamic topology self-organizing network.

2. A dynamic topology self-organizing network method based on environment perception and adaptive optimization according to claim 1, characterized in that: The physical environment parameters of the node include temperature gradient distribution data and obstacle distribution data. The specific method for real-time collection of the physical environment parameters of the node is as follows: The temperature changes in the area where the node is located are monitored with a sampling period of 10 seconds and an interval of 0.5°C to obtain temperature gradient distribution data; Infrared detection is used to identify and record the coordinates of obstacles with a resolution of 10 cm to obtain obstacle distribution data; The network status parameters of the node include the node remaining energy data and the communication link quality data. The specific method for real-time collection of the node network status parameters is as follows: Monitor the battery power of the node in real time with 1% accuracy to obtain the remaining energy data of the node; Evaluate the signal-to-noise ratio of the node. When the signal-to-noise ratio is greater than or equal to 15dB, determine that the node has reliable communication link quality and obtain communication link quality data; During the SNR evaluation process, the physical distance between nodes is calculated based on the signal propagation time; The dynamic interference parameters of a node include the channel congestion level and the signal attenuation coefficient. The specific method for real-time collection of the dynamic interference parameters of a node is as follows: The sampling period is 100ms, and the packet collision rate is used as the channel congestion level; The signal attenuation coefficient is dynamically calculated based on the received signal strength of the node, and its expression is: in Represents the signal attenuation coefficient of a single node; is the transmit power of the transmitting node; is the receiving power of the receiving node; is the distance between the transmitting node and the receiving node; the signal attenuation range is 0.5 dB / m ~3.0 dB / m.

3. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 2, characterized in that: The specific method of constructing a multidimensional quantization parameter matrix including an obstacle coverage matrix, an energy consumption distribution matrix, and a signal attenuation coefficient matrix includes the following steps: The physical environment parameters, network status parameters, and dynamic interference parameters of the nodes collected in real time are filtered for noise, and the target area is divided into 1m×1m grid cells. The physical environment parameters, network status parameters, and dynamic interference parameters of the nodes corresponding to each grid cell are obtained. Divide each grid cell into 100 identical sub-cells. Count the number of sub-cells containing obstacles based on the coordinates of the obstacles. Divide the number of sub-cells containing obstacles in a single grid cell by 100 to obtain the obstacle coverage rate in a single grid cell. The first i Rank j The obstacle coverage rate in the column grid cells is used as the first i Rank j The elements of the column are used to obtain the obstacle coverage matrix; The grid cells with node residual energy lower than 20% are recorded as high energy consumption areas, the grid cells with node residual energy between 20% and 60% are recorded as medium energy consumption areas, and the grid cells with node residual energy greater than 60% are recorded as low energy consumption areas. The energy consumption distribution values ​​of the low energy consumption area, medium energy consumption area, and high energy consumption area are set to 0, 0.5, and 1, respectively, to obtain the energy consumption distribution values ​​corresponding to each grid cell, and then the energy consumption distribution matrix is ​​obtained. The expression of the energy consumption distribution value in a single grid cell is: Indicates the i Rank j The energy consumption distribution value corresponding to the grid cell at the column position, that is, the energy consumption distribution matrix i Rank j The value of the element at the column position; The signal attenuation coefficient corresponding to each grid unit is obtained according to the signal attenuation coefficient of a single node, and the signal attenuation coefficient corresponding to each grid unit is used as an element of the signal attenuation coefficient matrix to obtain the signal attenuation coefficient matrix.

4. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 3, characterized in that: Based on the multi-dimensional quantization parameter matrix, the specific method of generating node weight values ​​by fusing path loss, communication delay and node energy consumption parameters includes the following steps: Calculate the obstacle attenuation factor between the transmitting node and the receiving node based on the obstacle coverage matrix The obstacle attenuation factor ranges from 1.0 to 5.

0. For every 10% increase in obstacle coverage, the obstacle attenuation factor increases by 0.

4. The path loss is calculated using the improved Friis free space model, and its expression is: in is the path loss; is the transmit power; is the antenna gain of the transmitting node; is the antenna gain of the receiving node; is the wavelength of the signal carrier; is pi; is the distance between the transmitting node and the receiving node; The communication delay is calculated based on the transmission distance and link utilization, and its expression is: in For communication delay; is the signal propagation speed; is the queue length; is bandwidth; is the link utilization; The node energy consumption is dynamically evaluated based on the remaining energy and energy consumption rate, and its expression is: in is the node energy consumption; is the current remaining energy of the node; is the initial energy of the node; is the amount of energy used; For usage time; Indicates the energy consumption rate; Generate node weight value, the expression is: in is the node weight; 、 and They are path loss weight, communication delay weight and node energy consumption weight respectively.

5. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 4, characterized in that: When the temperature in the area where the node is located is greater than or equal to 50℃, the The value of is reduced by 0.1, and The value of is increased by 0.15; When the obstacle coverage between the transmitting node and the receiving node is greater than 30%, each 10% increase will The value of increases by 0.2; when When , the queue delay weight of the corresponding node is increased by 50%; When the energy consumption rate of a node is higher than 30% of the average energy consumption rate of the entire network, the node energy consumption weight of the node will be reduced by 20%.

6. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 1, characterized in that: The specific method for calculating the optimal communication link priority of each node using a game theory strategy based on the multi-dimensional quantization parameter matrix and the node weight value includes the following steps: Perform preliminary link optimization with the goal of link load balancing through the Nash equilibrium strategy to obtain the initially optimized link; Based on the initially optimized links, the ant colony algorithm is combined with pheromone concentration and node weight value to perform global path optimization with communication efficiency as the goal, update the path selection probability, and obtain the selected path; Based on the selected path, the particle swarm algorithm is used to optimize network coverage and energy consumption balance, adjust the node positions, ensure balanced distribution of network resources, and complete the dynamic topology self-organizing network.

7. A dynamic topology self-organizing network method based on environment perception and adaptive optimization according to claim 6, characterized in that: The Nash equilibrium strategy is used to perform preliminary link optimization with link load balancing as the goal. The specific method for obtaining the preliminary optimized link is as follows: Dynamically balance node energy consumption weights through inter-node negotiation , so that the standard deviation of energy consumption distribution is less than or equal to 15%; When the standard deviation of the remaining energy of the nodes in the target area exceeds 15%, the energy consumption weight of the nodes corresponding to the high energy consumption area is increased. Reduce by 0.1 and make the node energy consumption weight corresponding to the nodes in the low energy consumption area Increase by 0.1; If the link load difference after preliminary optimization exceeds 20%, the dynamic priority allocation mechanism is triggered to keep the standard deviation of the utilization of each link within 10%.

8. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 7, characterized in that: The specific method of global path optimization with communication efficiency as the goal by combining pheromone concentration and node weight value with the ant colony algorithm includes the following steps: Taking the path as an individual of the ant colony, the ant colony algorithm is run with communication efficiency as the goal to obtain the pheromone concentration corresponding to the path; The weight of pheromone concentration is set to 40%, and the weight of node weight is set to 60%. The probability of path selection is calculated as follows: in For path The probability of selection; For path pheromone concentration; For path The node weight value of For path The node weight value of For path pheromone concentration; The path with the highest selection probability is selected as the selected path.

9. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 8, characterized in that: During the calculation of pheromone concentration, for links with a path loss not exceeding 80dB and a communication delay not exceeding 50ms, the pheromone concentration is increased by an additional 20%; the pheromone volatilization rate is set at 5% / minute.

10. The method of dynamic topology self-organizing network based on environment perception and adaptive optimization according to claim 6, characterized in that: The specific method of adjusting the node position by combining the node position and the node weight value with the network coverage and energy consumption balance as the optimization goal through the particle swarm algorithm is as follows: The node positions are used as particles in the particle swarm algorithm. The particle swarm algorithm is run with the network coverage greater than or equal to 85% and the energy consumption distribution standard deviation less than or equal to 15% as the optimization objectives to obtain the node positions.

Images