A method, system, device and medium for unmanned aerial vehicle cluster routing based on dynamic network diameter adaptation

Abstract

The application discloses a kind of unmanned aerial vehicle cluster routing method, system, equipment and medium based on dynamic network diameter self-adaption, which comprises: the position information of each node of unmanned aerial vehicle cluster is acquired in real time to calculate maximum network diameter, and routing parameter is dynamically adjusted;Real-time monitoring link received signal power and bit error rate to calculate normalized channel fading factor, adjust data transmission strategy;Real-time acquisition packet in the protocol stack processing delay and queuing delay of each hop node, calculate maximum transmission delay, adjust routing discovery and period;According to the change of current maximum network diameter, normalized channel fading factor, maximum transmission delay and packet loss rate, dynamically adjust network diameter time, so as to dynamically adjust the maintenance period of routing, path validity period and routing request retransmission interval;The system, equipment and medium are used to realize the method, and the dynamic network diameter self-adaptive routing algorithm proposed in the application improves the routing maintenance and self-healing ability of unmanned aerial vehicle cluster.

Description

Technical Field

[0001] This invention belongs to the field of UAV swarm communication and network technology, and particularly relates to a UAV swarm routing method, system, device and medium based on dynamic network diameter adaptive routing. Background Technology

[0002] In UAV swarm ad hoc networking scenarios, the main challenge in routing performance optimization lies in the high-frequency changes in dynamic topology and the vulnerability of communication link channels. UAV nodes typically move at speeds of 10 m / s to 20 m / s or even higher, with parameters such as heading and acceleration changing in real time. This high-speed movement causes instantaneous fluctuations in the physical distance between nodes to far exceed the adaptability threshold of traditional MANETs, ​​resulting in significant dynamic challenges for UAV swarms. For example, in a forest fire scenario, increasing the node spacing by 50 meters can lead to a sudden drop in received signal strength of 15 dB, triggering a link reconfiguration mechanism. Simultaneously, high-speed movement exacerbates multipath fading, increasing the probability of communication link interruption and further reducing communication reliability.

[0003] Traditional routing protocols have significant limitations in highly dynamic ad hoc network scenarios involving drone swarms. On the one hand, traditional protocols rely on fixed parameters (such as routing request intervals and path timeout thresholds), which can lead to a surge in redundant control messages when nodes move at high speeds. For example, the on-demand routing mechanisms of AODV (Ad hoc On-Demand Distance Vector Routing) and HWMP (Hybrid Wireless Mesh Protocol) require frequent path discovery processes, resulting in an excessively high proportion of control messages (K. Lei, Q. Zhang, J. Lou, B. Bai, and K. Xu, 'Securing ICN-based UAV ad hoc networks with blockchain,' IEEE Commun. Mag., vol. 57, no. 6, pp. 26–32, Jun. 2019.) (Wu Junsheng. Research and Application of HWMP Protocol Based on Wireless Mesh Networks [D]. Beijing Jiaotong University, 2016); while the OLSR (Optimized Link State Routing) protocol optimizes link state broadcasting by nodes, it still needs to periodically send Hello messages to maintain the neighbor table when the topology changes frequently, resulting in a waste of bandwidth resources (Zhang D, Yu FR, Yang R, et al.). al. Software-Defined Vehicular Networks With TrustManagement: A Deep Reinforcement Learning Approach[J]. Trans. Intell.Transport. Sys., 2022, 23(2): 1400-1414.). On the other hand, the preset path lifetime is difficult to match the actual network diameter of dynamic networks. Even after the path information has failed, transmission is still attempted, resulting in an increased packet loss rate and reduced overall network resource utilization efficiency. Summary of the Invention

[0004] To address the problems in existing technologies, the present invention aims to provide a method, system, device, and medium for UAV swarm routing based on dynamic network diameter adaptation. This method calculates the maximum network diameter by acquiring the location information of each node in the UAV swarm in real time, and dynamically adjusts routing parameters; monitors the received signal power and bit error rate of the link to calculate the normalized channel fading factor and adjust data transmission strategies; acquires the protocol stack processing delay and queuing delay of data packets at each hop node in real time, calculates the maximum transmission delay, and adjusts route discovery and the cycle; and dynamically adjusts the network diameter time based on changes in the current maximum network diameter, normalized channel fading factor, maximum transmission delay, and packet loss rate, thereby dynamically adjusting the route maintenance cycle, path validity period, and route request retransmission interval, improving the routing maintenance and self-healing capabilities of the UAV swarm.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A drone swarm routing method based on dynamic network diameter adaptation includes the following steps: Step 1: Obtain the location information of each node in the drone cluster in real time, calculate the maximum physical distance between any two nodes, and obtain the current maximum network diameter; Step 2: Monitor the received signal power and bit error rate of the link in real time, and calculate the normalized channel fading factor based on the ratio of received signal power to transmitted signal power; Step 3: Real-time acquisition of protocol stack processing delay and queuing delay of data packets at each hop node, accumulating the delay value hop by hop along the transmission path, and calculating the maximum transmission delay; Step 4: Based on the UAV swarm flight environment, calculate the current network diameter time according to the current maximum network diameter, normalized channel fading factor, maximum transmission delay, and packet loss rate; Step 5: Determine whether the current network diameter time is within the preset maximum and minimum range of the network diameter time; if yes, dynamically adjust the route maintenance cycle, path validity period, and route request retransmission interval according to the current network diameter time; if no, return to step 1.

[0006] In step 1, the calculation process for the current maximum network diameter is as follows: Step 1.1: Initialize global UAV node location information, including absolute position. and relative position Among them, drone nodes In the time slot The absolute position is represented as: drones Compared to drones position vector Represented as: ; Therefore, in time slots The relative position matrix R is represented as: in, Indicates the number of drones; Step 1.2: Based on the global UAV node location information, traverse all node pairs, calculate the 3D plane Euclidean distance, and obtain the maximum network diameter. The calculation formula is as follows: in, Indicates drone Compared to drones The position vector.

[0007] In step 2, the calculation process of the normalized channel fading factor is as follows: Step 2.1: Based on the path fading loss of the real channel, the total channel propagation loss can be obtained by superimposing the results. The expression is as follows: in, This represents the logarithmic distance path loss. Indicates multipath fading loss. Indicates shadow fading loss; Without considering transmitter and receiver losses, the received signal power is expressed as: in, For the transmitted signal power, For transmitter gain, For receiver gain; Step 2.2: Real-time acquisition of received signal power and bit error rate The channel fading factor is obtained. and effective bit error rate ,in, Indicates the weighting coefficient; Step 2.3: Based on the received signal power With transmitted signal power The ratio of the two factors yields the normalized channel fading factor, which is calculated using the following formula: .

[0008] In step 3, the calculation process for the maximum transmission delay is as follows: Step 3.1: Real-time statistics on the processing time of data packets in the protocol stack of each hop node. This includes routing table lookup latency. and data encryption latency ; Step 3.2: Calculate data encryption latency The calculation formula is as follows: in, Indicates the size of the plaintext data packet. Indicates hardware-accelerated throughput. Indicates the average CPU frequency; Step 3.3: Real-time statistics of the queuing delay of the IEEE 802.11s protocol. When the channel is busy, the node randomly selects a backoff counter value and sets the waiting time σ to decrease each time; therefore, the queuing delay is obtained. The calculation formula is as follows: in, Minimum competition window; This indicates the congestion adjustment weight, which is dynamically adjusted by the MAC layer. Specifically, for every additional collision, the weight is adjusted accordingly. One update iteration, that is ; Step 3.4: Accumulate the delay value hop-by-hop along the transmission path to obtain the maximum transmission delay. The calculation formula is as follows: in, This indicates the number of hops in network transmission between drones.

[0009] In step 4, the formula for calculating the network diameter time is as follows: in, For real-time packet loss rate, For environmental propagation coefficient, Normalized channel fading factor At the speed of light, This represents the maximum transmission delay.

[0010] The environmental propagation coefficient The expression is as follows: .

[0011] In step 5, the maintenance cycle, path validity period, and route request retransmission interval are dynamically adjusted based on the dynamic network diameter time, as follows: in, All are weighting coefficients. This represents the maximum number of retries.

[0012] This invention also provides a UAV swarm routing system based on dynamic network diameter adaptation, comprising: First calculation module: used to acquire the location information of each node in the drone cluster in real time, calculate the maximum physical distance between any two nodes, and obtain the current maximum network diameter; The second calculation module is used to monitor the received signal power and bit error rate of the link in real time, and calculate the normalized channel fading factor based on the ratio of received signal power to transmitted signal power. The third calculation module is used to obtain the protocol stack processing delay and MAC layer queuing delay of data packets at each hop node in real time, accumulate the delay value hop by hop along the transmission path, and calculate the maximum transmission delay. The fourth calculation module calculates the current network diameter time based on the real environment of UAV swarm flight, according to the current maximum network diameter, normalized channel fading factor, maximum transmission delay, and packet loss rate. Judgment module: Used to determine whether the current network diameter time is within the preset maximum and minimum range of network diameter time; if yes, dynamically adjust the route maintenance cycle, path validity period and route request retransmission interval according to the current network diameter time; if no, return to step 1.

[0013] The present invention also provides a drone swarm routing device based on dynamic network diameter adaptation, comprising: Memory: Used to store the computer program that implements the above-described UAV swarm routing method based on dynamic network diameter adaptation; Processor: Used to implement the above-described UAV swarm routing method based on dynamic network diameter adaptation when executing the computer program.

[0014] The present invention also provides a computer-readable storage medium storing a computer program that, when executed by a processor, enables the implementation of the above-described UAV swarm routing method based on dynamic network diameter adaptation.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The channel fading factor defined in this invention does not evaluate channel state using only a single indicator such as received signal strength, signal-to-noise ratio, or bit error rate. Instead, it simultaneously introduces the ratio of received signal power to transmitted signal power at the physical layer, as well as the real-time bit error rate fed back from the link layer, to perform cross-layer fusion characterization of channel state. The ratio of received signal power to transmitted signal power reflects signal energy attenuation caused by path loss, multipath fading, shadowing effects, and obstacle obstruction; the bit error rate reflects the actual impact of noise, interference, and link fluctuations on the ability to transmit data correctly. By setting an exponential penalty term for the bit error rate, the distinction between high-error-rate links and normal links can be enhanced, preventing links with high received signal power but poor actual transmission reliability from being incorrectly evaluated as high-quality links. Therefore, this invention is not limited to proposing a new mathematical parameter, but rather aims to convert channel state information from different protocol layers and with different dimensions into a unified, normalized, and real-time updatable channel characterization quantity. This channel fading factor can reflect both the energy attenuation of wireless signals and the reliability of the data actually carried by the link. Compared with judging solely based on node distance or the power of a single received signal, it can more comprehensively and accurately describe the actual channel state of UAV swarms in high-speed movement, multipath propagation, and complex obstruction environments, and provide a directly usable quantitative basis for the dynamic adjustment of network layer routing parameters.

[0016] 2. This invention incorporates a normalized channel fading factor into the calculation of network diameter time, subjecting it to the combined constraints of network spatial scale and real-time channel conditions. This corrects the ideal propagation time, which is based solely on geometric distance, into an effective network traversal time that reflects link transmission success rate, bit error rate, and retransmission impact. In other words, even if the maximum physical distance of the UAV swarm does not change significantly, when obstruction, interference, or multipath fading degrades the channel conditions, the system can still promptly correct the network diameter time based on changes in the channel fading factor, preventing routing parameters from continuing to operate under ideal channel conditions.

[0017] 3. This invention proposes a Dynamic Network Diameter Adaptive Routing Algorithm (DD-HWMP). Its core optimization lies in dynamically adjusting the network diameter time based on changes in network topology (current maximum network diameter, normalized channel fading factor, maximum transmission delay) and packet loss rate. This dynamically adjusts the route maintenance cycle, path validity period, and route request retransmission interval to adapt to the highly dynamic network environment of UAV swarm ad hoc networks, thereby reducing packet loss rate and increasing network throughput.

[0018] In summary, the Dynamic Network Diameter Adaptive Routing Algorithm (DD-HWMP) proposed in this invention significantly outperforms traditional protocols (such as HWMP and AODV) in key metrics such as average end-to-end latency, packet loss rate, and network throughput, and exhibits stronger adaptability and stability, especially in highly dynamic scenarios. Attached Figure Description

[0019] Figure 1 This invention provides a UAV swarm routing method based on dynamic network diameter adaptation.

[0020] Figure 2 This is a schematic diagram of drone swarm communication in a forest fire scenario using the method of the present invention.

[0021] Figure 3 This is a graph showing the change of network diameter over time as a function of simulation time in this embodiment.

[0022] Figure 4 This is a comparison chart of the average end-to-end latency of three routing protocols: DD-HWMP, HWMP, and AODV.

[0023] Figure 5 This is a comparison chart of packet loss rates for three routing protocols: DD-HWMP, HWMP, and AODV.

[0024] Figure 6 This is a comparison chart of network throughput for three routing protocols: DD-HWMP, HWMP, and AODV.

[0025] Figure 7 This is a comparison chart of the average end-to-end delay of DD-HWMP and HWMP under different node movement speeds.

[0026] Figure 8 This is a comparison chart of packet loss rates between DD-HWMP and HWMP at different node movement speeds.

[0027] Figure 9 This is a comparison chart of the average network throughput of DD-HWMP and HWMP under different node movement speeds. Detailed Implementation

[0028] The present invention will be further described below with reference to the accompanying drawings and embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art should fall within the protection scope of the present invention.

[0029] This invention focuses on the core parameter of Network Diameter Time (NDT). NDT is defined as the estimated time for a data packet to traverse the maximum diameter of the network, denoted as T. dIn the HWMP routing protocol, NDT determines the route maintenance cycle T. PREQ Path validity period T active and route request retransmission interval T retry Traditional HWMP protocols employ preset fixed NDT (default 100ms) and path lifetime (default 500ms), failing to consider the dynamic coupling relationship between node movement speed, channel quality, and topology changes. This design leads to a mismatch between path lifetime and topology deformation rate in high-speed UAV scenarios, causing frequent routing failures. Simultaneously, in low-dynamic scenarios, redundant routing maintenance messages are continuously triggered. In medium-density networks, the control overhead of the HWMP protocol accounts for 30%-50%, with 60% of PREQ / PREP (Path Request / Path Response) frames originating from unnecessary periodic broadcasts, violating the core energy efficiency requirements of UAV swarms. If the NDT is less than the actual network diameter time, it may be impossible to traverse all nodes in the network, failing to establish complete routing information, resulting in high packet loss rates and generating excessive control packets, leading to invalid PREQ flooding and increased network overhead. Conversely, if the NDT is set too large, routing response is too slow in high-dynamic network environments, resulting in low network efficiency and excessive transmission latency. Therefore, the network diameter time needs to be controlled within a reasonable range based on the actual network topology.

[0030] like Figure 1 As shown, this invention proposes a UAV swarm routing method based on dynamic network diameter adaptation, namely the Dynamic Diameter-Hybrid Wireless Mesh Protocol (DD-HWMP) algorithm. This method controls the network diameter time within a reasonable range according to the actual network topology, thereby dynamically adjusting the routing maintenance cycle T. PREQ Path validity period T active and route request retransmission interval T retry Specifically, it includes the following steps: Step 1, Network Diameter Dynamic Sensing: Real-time acquisition of the location information of each node in the drone swarm, calculation of the maximum physical distance between any two nodes, calculation of the current maximum network diameter, and dynamic adjustment of routing parameters based on the maximum network diameter; Specifically, a sliding window maximum value detection strategy is adopted to update the maximum distance of the global network diameter in real time. The UAVs are uniformly distributed in a three-dimensional space of X×Y×Z, with a total of S UAVs. The flight mission in this invention employs a Gaussian-Markov random algorithm; the current speed and direction of the UAV are determined by the previous state and a Gaussian random variable.

[0031] Step 1.1: Initialize global UAV node location information, including absolute position. and relative position Among them, drone nodes In the time slot The absolute position is represented as: drones Compared to drones position vector Represented as: Therefore, the relative position vectors between drones are calculated from their absolute positions, within the time slot. The relative position matrix R is represented as: .

[0032] Step 1.2: Based on the global UAV node position information, traverse all node pairs, calculate the 3D plane Euclidean distance, and achieve dynamic matching of node pairs through a double loop to obtain the maximum network diameter. The calculation formula is as follows: .

[0033] This achieves dynamic sensing of network diameter.

[0034] Step 2, Channel Fading Factor Calculation: Monitor the received signal power and bit error rate of the link in real time, calculate the normalized channel fading factor based on the ratio of received signal power to transmitted signal power, and adjust the data transmission strategy according to the channel fading factor. Step 2.1: Based on the path fading loss of the real channel, the total channel propagation loss can be obtained by superimposing the results. The expression is as follows: in, This represents the logarithmic distance path loss. Indicates multipath fading loss. This indicates the loss due to shadow fading.

[0035] Without considering transmitter and receiver losses, the received signal power is expressed as: in, For the transmitted signal power, For transmitter gain, This is the receiver gain.

[0036] Step 2.2: Real-time statistics of received signal power from the MAC layer (Medium Access Control layer). The channel fading factor is obtained. .

[0037] Error rate is statistically analyzed via link layer feedback. An exponential decay model is used to suppress high-error-rate links, achieving an effective bit error rate. The calculation formula is: in, This represents the weighting coefficient, with a value of 0.5. Step 2.3: Based on the received signal power With transmitted signal power The ratio of the two factors yields the normalized channel fading factor, which is calculated using the following formula: Step 3, Transmission delay compensation: Real-time acquisition of the protocol stack processing delay and MAC layer queuing delay of data packets at each hop node, accumulating the delay value hop by hop along the transmission path, calculating the maximum transmission delay, and adjusting the route discovery and maintenance cycle based on the maximum transmission delay.

[0038] Step 3.1: Analyze the real-time data processing time of data packets in the protocol stack at each hop node from the MAC layer. This includes routing table lookup latency. and data encryption latency The latency for routing table lookups is calculated using a CPU clock cycle counter. ; In this embodiment, the typical value of the routing table lookup latency is obtained based on the determined CPU clock cycle counter. ; Step 3.2: Calculate data encryption latency The calculation formula is as follows: in, Indicates the size of the plaintext data packet, in bytes; This represents the hardware-accelerated throughput, which is obtained using the AES-128 algorithm. Indicates the average CPU frequency; In this embodiment, the selected hardware is a Ryzen 7 5800H single-core AES-128-CBC, with an encryption performance of approximately 3.5 cycles / byte.

[0039] Step 3.3: Real-time statistics of the MAC layer queuing delay of the IEEE 802.11s protocol are obtained from the MAC layer, representing the backoff and queue management delay of the contention window. When the channel is busy, the node randomly selects a backoff counter value and sets a waiting time σ to decrease each time; therefore, the MAC layer queuing delay is obtained. The calculation formula is as follows: in, The minimum contention window is set to 16. This indicates the congestion adjustment weight, which is dynamically adjusted by the MAC layer. Specifically, for every additional collision, the weight is adjusted accordingly. One update iteration, that is .

[0040] In this embodiment, the waiting time σ = 9 μs.

[0041] Step 3.4: Accumulate the delay value hop-by-hop along the transmission path to obtain the maximum transmission delay. The calculation formula is as follows: in, This indicates the number of hops in network transmission between drones; Step 4: Dynamic adjustment of network diameter time: The network diameter time is dynamically adjusted according to the packet loss rate. When the packet loss rate increases, the path validity period and route request retransmission interval are extended, and the frequency of sending invalid route discovery packets is reduced. By dynamically adjusting the network diameter time, route maintenance and self-healing capabilities are optimized.

[0042] A packet loss rate redundancy term is introduced to dynamically optimize the grid diameter time. For every 1% increase in packet loss rate, the network diameter time is reduced. An additional 2% increase is achieved by extending the routing maintenance cycle. and path validity period To suppress routing oscillations, the current network diameter time is calculated based on the current maximum network diameter, normalized channel fading factor, maximum transmission delay, and packet loss rate. The calculation formula is as follows: in, For real-time packet loss rate, Normalized channel fading factor At the speed of light, This represents the maximum transmission delay. This refers to the environmental propagation coefficient. The environmental propagation coefficient is used to simulate the real environment of drone swarm flight. The expression is as follows: Step 5: Determine if the current network diameter time is within the preset maximum value of the network diameter time. and minimum value Within the range, that is If yes, then dynamically adjust the route maintenance cycle, path validity period, and route request retransmission interval based on the current network diameter time; if no, return to step 1. Based on optimized dynamic network diameter time To dynamically adjust the maintenance cycle of the routing. Path validity period and route request retransmission interval To improve the path maintenance and self-healing capabilities of the routing algorithm; dynamic adjustments are made as follows: in, All are weighting coefficients, default. , ; This represents the maximum number of retries.

[0043] like Figure 2 The diagram illustrates a scenario where the method of this invention is applied. Assume a forest fire occurs. When the fire breaks out, it spreads rapidly, and traditional manual reconnaissance methods cannot obtain real-time information about the fire scene in a timely manner. Swarm UAVs, with their rapid deployment and flexible mobility, can arrive at the scene immediately after the fire breaks out to conduct real-time reconnaissance and data collection. Swarm UAVs need to form a coverage network over the fire scene, transmitting the collected images and data back to the command center via multi-hop routing, providing crucial support for firefighting decisions. The UAVs perform data collection tasks within a designated area and transmit the collected data to the base station. All UAV nodes are isomorphic and fly in formation to their designated areas. During operation, the nodes move autonomously according to the UAV's movement model. The UAV nodes are evenly distributed across the fire scene according to the network, with each node representing one UAV. The nodes are interconnected via wireless links. This deployment method can effectively cover a large area while ensuring that the communication distance between nodes remains within a reasonable range. During the initial formation flight, the horizontal and vertical distances between nodes remain consistent to ensure the stability of the initial coverage and communication links. After reaching the designated location, the UAVs move randomly within the mission area to simulate mission execution.

[0044] The technical effects of the present invention will be described in detail below with reference to simulation.

[0045] like Figure 3 As shown, it can be seen that the network topology changes continuously with time. During the 10-20s period, the network diameter time decreases, and the communication range of the entire drone swarm is in a shrinking state. After 20s, the network diameter time gradually increases, the communication range of the drone swarm gradually increases, and the relative distance between drones also gradually increases.

[0046] To evaluate the performance of the proposed DD-HWMP routing algorithm in a drone swarm environment, a simulation program was designed and its results were discussed to highlight the advantages of the proposed scheme. The experiment was conducted under the following conditions: (1) When the node speed was 10 m / s and the simulation time was 100 s, the average end-to-end delay of HWMP, AODV, and the proposed DD-HWMP was compared. (2) When the node speed was 10 m / s and the simulation time was 100 s, the network packet loss rate of HWMP, AODV, and the proposed DD-HWMP was compared. (3) When the node speed was 10 m / s and the simulation time was 100 s, the network throughput of HWMP, AODV, and the proposed DD-HWMP was compared. (4) With the simulation time set to 100 s and the average moving speed of the nodes changed, the average end-to-end delay of HWMP and the proposed DD-HWMP was compared. (5) Set the simulation time to 100s, change the average moving speed of the nodes, and conduct a comparative performance analysis of the network packet loss rate between HWMP and the proposed DD-HWMP. (6) Set the simulation time to 100s, change the average moving speed of the nodes, and conduct a comparative performance analysis of the network throughput between HWMP and the proposed DD-HWMP.

[0047] like Figure 4 As shown, this paper compares and analyzes the average end-to-end latency performance of DD-HWMP, HWMP, and AODV protocols. It represents the average latency from network packet generation to accurate reception. In the first 10-30 seconds, the average end-to-end latency of HWMP, AODV, and DD-HWMP is not significantly different, all between 10-20ms. However, as time progresses, drone nodes move continuously, the local topology changes rapidly, and routes switch constantly. The average end-to-end latency of HWMP and AODV increases rapidly, even reaching 180ms in the later stages of the simulation. In contrast, the average end-to-end latency of DD-HWMP remains very stable, not exceeding 25ms at the end of the simulation. This is because DD-HWMP can adjust the latency based on the real-time network diameter time T. d Adjusting the PREQ packet sending interval allows the route discovery and route maintenance cycles to adapt to the current network topology, improving route discovery speed, optimizing path quality, and reducing latency. This demonstrates that the DD-HWMP routing algorithm proposed in this invention can better adapt to changes in network topology.

[0048] like Figure 5The diagram shows a performance comparison of network packet loss rates between DD-HWMP and HWMP, and between HWMP and AODV protocols. It can be seen that AODV's packet loss rate is higher than both HWMP and DD-HWMP throughout the simulation, which is related to AODV's path selection principle of minimizing hops. During the 10-37s period, DD-HWMP's average packet loss rate is higher than HWMP's; however, as the simulation progresses, DD-HWMP's packet loss rate becomes lower than HWMP's. This is because DD-HWMP, under high packet loss conditions, increases the network diameter time T. d The longer retransmission interval reduces invalid PREQ flooding; the extended path validity period prioritizes the use of existing paths, reduces PREQ transmission frequency, and lowers control overhead. As a result, during periods of rapid network topology change, the average packet loss rate of DD-HWMP is lower than that of HWMP, indicating that the DD-HWMP routing algorithm proposed in this invention has a stronger routing self-healing capability.

[0049] like Figure 6 As shown, the performance comparison analysis of network throughput of the DD-HWMP protocol of this invention with HWMP and AODV protocols is presented. It can be seen that after the simulation starts, the network throughput of HWMP, AODV, and DD-HWMP all increase rapidly. This is because at the beginning of the simulation, the entire network is still in the route establishment phase, so the network throughput is low; however, as routes are gradually established, the network throughput increases rapidly. DD-HWMP has the highest network throughput compared to HWMP and AODV in the overall simulation. As the simulation progresses, the network throughput of HWMP and DD-HWMP tends to stabilize, while the network throughput of AODV gradually decreases after 60 seconds of simulation time. This is related to... Figure 5 This is related to the significant increase in packet loss rate after 60 seconds in AODV.

[0050] Figure 7 , Figure 8 and Figure 9 The graphs show a comparison of the average end-to-end latency, packet loss rate, and average network throughput of DD-HWMP and HWMP at different node speeds. Figure 7As shown, with increasing node movement speed, network topology changes more rapidly, leading to more unstable links and limited effectiveness of routing information. For HWMP, increased node movement speed has a significant impact on average end-to-end latency because the rapid changes in network topology mean that fixed path discovery cycles and path lifetimes cannot keep up with the pace of change, resulting in significant time consumption for path retransmissions and a substantial increase in end-to-end latency. In contrast, for DD-HWMP, average end-to-end latency does not change significantly with node speed because DD-HWMP can dynamically adjust the network diameter time based on network topology changes, thereby adjusting the path discovery and maintenance cycle, path lifetime, and route retransmission interval. This significantly reduces invalid route discovery packets caused by rapid network topology changes, thus DD-HWMP's average end-to-end latency does not change much with increasing node movement speed. Frequent route failures lead to an increase in packet loss rate, such as... Figure 8 As shown, although the average packet loss rate of both HWMP and DD-HWMP increases with the increase of node movement speed, at the same node movement speed, the packet loss rate of DD-HWMP is consistently about 5% lower than that of HWMP. Simulation results demonstrate that the DD-HWMP routing algorithm proposed in this invention is more adaptable to high-speed node movement scenarios than HWMP. Figure 9 As shown, in a highly dynamic environment, the DD-HWMP routing algorithm proposed in this invention has higher network throughput than HWMP, indicating that the DD-HWMP routing algorithm has stronger topology adaptability and link stability.

Claims

1. A UAV swarm routing method based on dynamic network diameter adaptation, characterized in that, Includes the following steps: Step 1: Obtain the location information of each node in the drone cluster in real time, calculate the maximum physical distance between any two nodes, and obtain the current maximum network diameter; Step 2: Monitor the received signal power and bit error rate of the link in real time, and calculate the normalized channel fading factor based on the ratio of received signal power to transmitted signal power; Step 3: Real-time acquisition of protocol stack processing delay and queuing delay of data packets at each hop node, accumulating the delay value hop by hop along the transmission path, and calculating the maximum transmission delay; Step 4: Based on the UAV swarm flight environment, calculate the current network diameter time according to the current maximum network diameter, normalized channel fading factor, maximum transmission delay, and packet loss rate; Step 5: Determine whether the current network diameter time is within the preset maximum and minimum range of the network diameter time; if yes, dynamically adjust the route maintenance cycle, path validity period, and route request retransmission interval according to the current network diameter time; if no, return to step 1.

2. The UAV swarm routing method according to claim 1, characterized in that: In step 1, the calculation process for the current maximum network diameter is as follows: Step 1.1: Initialize global UAV node location information, including absolute position. and relative position Among them, drone nodes In the time slot The absolute position is represented as: drones Compared to drones position vector Represented as: ； Therefore, in time slots The relative position matrix R is represented as: in, Indicates the number of drones; Step 1.2: Based on the global UAV node location information, traverse all node pairs, calculate the 3D plane Euclidean distance, and obtain the maximum network diameter. The calculation formula is as follows: in, Indicates drone Compared to drones The position vector.

3. The UAV swarm routing method according to claim 1, characterized in that: In step 2, the calculation process of the normalized channel fading factor is as follows: Step 2.1: Based on the path fading loss of the real channel, the total channel propagation loss can be obtained by superimposing the results. The expression is as follows: in, This represents the logarithmic distance path loss. This indicates multipath fading loss. Indicates shadow fading loss; Without considering transmitter and receiver losses, the received signal power is expressed as: in, For the transmitted signal power, For transmitter gain, For receiver gain; Step 2.2: Real-time acquisition of received signal power and bit error rate The channel fading factor is obtained. and effective bit error rate ,in, Indicates the weighting coefficient; Step 2.3: Based on the received signal power With transmitted signal power The ratio of the two factors yields the normalized channel fading factor, which is calculated using the following formula: 。 4. The UAV swarm routing method according to claim 1, characterized in that: In step 3, the calculation process for the maximum transmission delay is as follows: Step 3.1: Real-time statistics on the processing time of data packets in the protocol stack of each hop node. This includes routing table lookup latency. and data encryption latency ; Step 3.2: Calculate data encryption latency The calculation formula is as follows: in, Indicates the size of the plaintext data packet. Indicates hardware acceleration throughput. Indicates the average CPU frequency; Step 3.3: Real-time statistics of the queuing delay of the IEEE 802.11s protocol. When the channel is busy, the node randomly selects a backoff counter value and sets the waiting time σ to decrease each time; therefore, the queuing delay is obtained. The calculation formula is as follows: in, Minimum competition window; This indicates the congestion adjustment weight, which is dynamically adjusted by the MAC layer. Specifically, for every additional collision, the weight is adjusted accordingly. One update iteration, that is ; Step 3.4: Accumulate the delay value hop-by-hop along the transmission path to obtain the maximum transmission delay. The calculation formula is as follows: in, This indicates the number of hops in network transmission between drones.

5. The UAV swarm routing method according to claim 1, characterized in that: In step 4, the formula for calculating the network diameter time is as follows: in, For real-time packet loss rate, For environmental propagation coefficient, Normalized channel fading factor At the speed of light, This represents the maximum transmission delay.

6. The UAV swarm routing method according to claim 5, characterized in that: The environmental propagation coefficient The expression is as follows: 。 7. The UAV swarm routing method according to claim 1, characterized in that: In step 5, the maintenance cycle, path validity period, and route request retransmission interval are dynamically adjusted based on the dynamic network diameter time, as follows: in, All are weighting coefficients. This represents the maximum number of retries.

8. A UAV swarm routing system based on dynamic network diameter adaptation, characterized in that, include: First calculation module: used to acquire the location information of each node in the drone cluster in real time, calculate the maximum physical distance between any two nodes, and obtain the current maximum network diameter; The second calculation module is used to monitor the received signal power and bit error rate of the link in real time, and calculate the normalized channel fading factor based on the ratio of received signal power to transmitted signal power. The third calculation module is used to obtain the protocol stack processing delay and queuing delay of data packets at each hop node in real time, accumulate the delay value hop by hop along the transmission path, and calculate the maximum transmission delay. The fourth calculation module calculates the current network diameter time based on the real environment of UAV swarm flight, according to the current maximum network diameter, normalized channel fading factor, maximum transmission delay, and packet loss rate. Judgment module: Used to determine whether the current network diameter time is within the preset maximum and minimum range of network diameter time; if yes, dynamically adjust the route maintenance cycle, path validity period and route request retransmission interval according to the current network diameter time; if no, return to step 1.

9. A drone swarm routing device based on dynamic network diameter adaptation, characterized in that, include: Memory: for storing a computer program that implements the UAV swarm routing method based on dynamic network diameter adaptation as described in any one of claims 1-7; Processor: Used to implement the UAV swarm routing method based on dynamic network diameter adaptation as described in any one of claims 1-7 when executing the computer program.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that, when executed by a processor, enables the implementation of the UAV swarm routing method based on dynamic network diameter adaptation as described in any one of claims 1-7.

Images