Unmanned aerial vehicle network routing method under spectrum denial based on low-orbit satellite cooperation

Abstract

The application discloses a spectrum denial under unmanned aerial vehicle network routing method based on low-orbit satellite cooperation, the method aims at unmanned aerial vehicle communication problem under spectrum interference, introduces low-orbit satellite as the auxiliary communication node of unmanned aerial vehicle network, and uses network controller to formulate a routing scheme.Specifically, the network controller first formulates an initial routing scheme for all data streams based on global network information, including calculating transmission path and allocating network bandwidth, and evaluating the performance of the routing scheme, then using the simulated annealing framework, on the basis of the initial routing scheme, iteratively optimizing to explore better solutions.The application has the advantages that it avoids the problem of too long solving time of the precise algorithm, and avoids the problem that the traditional path search algorithm is easy to fall into local optimal solution, and can obtain approximate solution in a short time.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of unmanned aerial vehicle network routing under spectrum denial, in particular to a method for unmanned aerial vehicle network routing under spectrum denial based on low-orbit satellite cooperation. BACKGROUND

[0002] Unmanned aerial vehicles have been widely used in various industries due to their high flexibility, rapid response, and strong environmental adaptability. Unmanned aerial vehicle networks are a special type of wireless ad hoc network, where multiple unmanned aerial vehicles interact with each other through communication to achieve group advantages and complete more functions. In the military field, unmanned aerial vehicles can perform tactical reconnaissance, situation awareness, long-range strikes, and communication relay, and their position in the weapon system is increasingly prominent. Spectrum interference is a threat that unmanned aerial vehicle networks will inevitably encounter in future battlefields. Interference signals can degrade the quality of unmanned aerial vehicle communication links, cause data transmission delays or interruptions, or affect the navigation system of unmanned aerial vehicles, making them lose positioning ability, deviate from the planned route, and reduce the battlefield effectiveness of unmanned aerial vehicles, affecting intelligence acquisition and combat operations.

[0003] One of the main strategies to deal with interference is cognitive radio technology. Unmanned aerial vehicles dynamically perceive the radio environment and adjust communication parameters to avoid interference sources according to environmental changes. The core of this strategy is the adaptive ability, which can ensure the stability of the communication link in complex electromagnetic environments. Another strategy is trajectory planning technology, which optimizes the flight path of unmanned aerial vehicles based on geographic information and real-time environmental data to avoid known or potential interference areas and improve mission success rate. For unmanned aerial vehicle clusters, nodes that cannot communicate can be discarded, and the remaining nodes can be reorganized to form a formation, taking advantage of the autonomy and cooperation of unmanned aerial vehicles to quickly recover the network. The above two strategies have certain limitations in practical application: 1) Cognitive radio technology relies on efficient spectrum sensing algorithms and fast spectrum switching capabilities, which require high hardware and computing capabilities of unmanned aerial vehicles. At the same time, due to the limited spectrum resources, it may still be limited in a strong interference environment, making it difficult to find an undisturbed frequency band for communication. 2) In the case of frequent enemy interference changes, unmanned aerial vehicles may not be able to adjust their positions in time, and the effectiveness of trajectory planning will be limited. In addition, moving unmanned aerial vehicles to restore network connectivity may conflict with the original mission and have a negative impact on combat plans.

[0004] The rapid development of low-orbit satellites brings new development opportunities for unmanned aerial vehicles. Low-orbit satellites have the advantages of low link loss, small transmission delay, etc., and perform outstandingly in terms of anti-interference capability and communication stability, and can assist unmanned aerial vehicle networks to communicate under spectrum interference. Therefore, the invention introduces low-orbit satellites as auxiliary communication nodes of unmanned aerial vehicle networks, designs a routing method for unmanned aerial vehicle networks under spectrum denial based on low-orbit satellite cooperation, plans the communication path through the simulated annealing algorithm and reasonably allocates network bandwidth resources, effectively improving the system communication quality. In particular, for large-scale unmanned aerial vehicle networks, this method not only avoids the problem of too long solving time of the exact algorithm, but also avoids the problem that the traditional path search algorithm is easy to fall into a local optimal solution, and can obtain an approximate solution in a short time, which has practical significance and good application scenarios. SUMMARY

[0005] The purpose of the invention is to overcome the shortcomings of the prior art. The invention uses low-orbit satellites as auxiliary communication nodes of unmanned aerial vehicle networks, provides a routing method for unmanned aerial vehicle networks under spectrum denial based on low-orbit satellite cooperation, and realizes reliable routing between unmanned aerial vehicles under spectrum interference and improves communication quality.

[0006] Technical scheme: The routing method for unmanned aerial vehicle networks under spectrum denial based on low-orbit satellite cooperation, characterized in that the unmanned aerial vehicle network comprises a network controller, a low-orbit satellite, a plurality of master unmanned aerial vehicles and a plurality of slave unmanned aerial vehicles, each master unmanned aerial vehicle is provided with a satellite communication module and can perform bidirectional communication with the low-orbit satellite, all unmanned aerial vehicles can communicate with neighbor unmanned aerial vehicles under non-interference, the network controller can control all unmanned aerial vehicles and, after being interfered by spectrum, decides the routing of the entire network based on global network information and formulates a transmission scheme for all data streams in the network, specifically comprising the following steps:

[0007] (1) The network controller obtains global network information: the network controller interacts with the unmanned aerial vehicles to obtain information such as unmanned aerial vehicle positions, interference conditions and data streams to be transmitted in the network;

[0008] (2) Formulate an initial routing scheme: based on the global network information, the network controller first calculates the transmission path for all data streams, then allocates network bandwidth, and obtains the initial routing scheme of the data streams;

[0009] (3) Evaluate the performance of the initial routing scheme: the network controller respectively evaluates the delay performance and network bandwidth performance of all data streams, calculates the routing performance of the data streams in combination with a weight adjustment parameter, and further evaluates the comprehensive routing performance of the initial routing scheme;

[0010] (4) Optimize the routing scheme: based on the simulated annealing framework, the network controller iteratively optimizes the path selection and network bandwidth allocation of the initial routing scheme until the termination condition is met, and obtains the final routing scheme.

[0011] Further, the detailed steps of the step (2) are as follows:

[0012] (2.1) For all data flows, use Dijkstra algorithm to calculate the shortest delay path from the start point to the end point as the transmission path.

[0013] (2.2) Since there can be several data flows passing through the same node, and the network bandwidth demand of different data flows varies, the network controller needs to allocate network bandwidth to the data flows. For any two nodes u and v, define C(u, v) as the upper limit of network bandwidth between u and v. When there is only one data flow from node u to node v, if C(u, v) is greater than the demand of the data flow, the network bandwidth allocated to the data flow is equal to its demand, otherwise C(u, v) is allocated to the data flow. When there are multiple data flows from node u to node v, if C(u, v) is greater than the sum of the network bandwidth demands of these data flows, the network bandwidth allocated to each data flow is equal to its network bandwidth demand; otherwise, C(u, v) is proportionally allocated to these data flows according to their network bandwidth demands. After calculating the network bandwidth allocation scheme for all nodes, the actual available network bandwidth of a data flow is equal to the minimum value of the network bandwidth allocated to it by all nodes on its path.

[0014] Further, the detailed steps of the step (3) are as follows:

[0015] (3.1) For any data flow F in the network i , evaluate its delay performance S(h i ) and express it as:

[0016]

[0017] where h i represents the end-to-end delay of the data flow F i , D i represents the expected delay of F i , and exp represents the exponential function with base e.

[0018] (3.2) For any data flow F in the network i , evaluate its network bandwidth performance S(f i ) and express it as:

[0019]

[0020] where R i represents the data flow F inetwork bandwidth requirement.

[0021] (3.3) Since different data streams have different requirements on latency and network bandwidth, for any data stream F i in the network, its routing performance S(f i ) is equal to the weighted sum of the latency performance S(h i ) and the network bandwidth performance S(f i ), expressed as:

[0022] S i = λ i S(h i )+(1-λ i )S(f i )

[0023] where λ i represents the weight adjustment parameter of the data stream F i .

[0024] (3.4) Based on the routing performance of each data stream, the comprehensive routing performance S of the initial routing scheme is evaluated, which is the sum of the routing performance of all data streams.

[0025] Further, the detailed steps of the routing optimization in step (4) are as follows:

[0026] (4.1) Set the initial temperature, termination temperature, cooling coefficient, and maximum number of cycles of simulated annealing, and set the current temperature equal to the initial temperature;

[0027] (4.2) Based on the current routing scheme, generate a new routing scheme according to the following steps:

[0028] (4.2.1) Randomly select a data stream F m , whose transmission path is p m , and the starting point and the ending point are s m and t m respectively. Construct an intermediate node set containing all nodes that meet the following conditions: the node is connected to s m ; the node is connected to t m ; and the node is not on p m .

[0029] (4.2.2) Randomly select a node n from the intermediate node set, and use Dijkstra algorithm to calculate the shortest latency path p m from s n1 to node n and the shortest latency path p m from node n to t n2 , and combine p n1 and p n2 to obtain a new path

[0030] (4.2.3) Re-allocate network bandwidth for all data flows according to the network bandwidth allocation scheme in step (2.2) to obtain a new routing scheme.

[0031] (4.3) Evaluate the performance of the new routing scheme according to the performance evaluation method in step (3), and use the Metropolis acceptance criterion to determine whether to accept the new routing scheme. If the comprehensive routing performance of the new routing scheme is greater than the comprehensive routing performance of the current routing scheme, the new routing scheme is accepted; otherwise, accept the new routing scheme with a probability p accept the new routing scheme is accepted:

[0032]

[0033] wherein exp represents an exponential function with the natural constant e as the base, S new represents the comprehensive routing performance of the new routing scheme, S temp represents the comprehensive routing performance of the current routing scheme, and T represents the current temperature of the simulated annealing system.

[0034] (4.4) If the number of cycles at the current temperature reaches the maximum number of cycles, multiply the current temperature by the cooling coefficient to update the value of the current temperature, and jump to step (4.5), otherwise return to step (4.2);

[0035] (4.5) If the current temperature is less than the termination temperature, end the loop and output the current routing scheme as the optimal routing scheme, otherwise return to step (4.2).

[0036] Beneficial effects: The low-orbit satellite cooperative-based spectrum denial unmanned aerial vehicle network routing method designed by the application uses low-orbit satellites as auxiliary communication nodes of unmanned aerial vehicle networks, effectively improves system communication quality by planning communication paths and reasonably allocating network bandwidth resources. In particular, for large-scale unmanned aerial vehicle networks, this method not only avoids the problem of too long solving time of accurate algorithms, but also avoids the problem that traditional path search algorithms are prone to fall into local optimal solutions, and can obtain an approximate solution in a short time. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 is a flowchart of the application.

[0038] Figure 2 is a comparison chart of comprehensive routing performance simulation results of various algorithms in a low-load scenario.

[0039] Figure 3 is a comparison chart of average delay performance simulation results of various algorithms in a low-load scenario.

[0040] Figure 4Figure 1 is a comparison chart of simulation results of average network bandwidth performance of each algorithm in a low-load scenario.

[0041] Figure 5 Figure 4 is a comparison chart of simulation results of comprehensive routing performance of each algorithm in a high-load scenario.

[0042] Figure 6 Figure 5 is a comparison chart of simulation results of average time delay performance of each algorithm in a high-load scenario.

[0043] Figure 7 Figure 6 is a comparison chart of simulation results of average network bandwidth performance of each algorithm in a high-load scenario. DETAILED DESCRIPTION

[0044] The application will be further described in detail below in combination with the drawings and specific embodiments.

[0045] The UAV network routing method based on low-orbit satellite cooperation under spectrum denial includes a network controller, a low-orbit satellite, and a UAV network. The UAV network is composed of multiple master UAVs and slave UAVs that can move freely. Each master UAV is equipped with a satellite communication module and can perform bidirectional communication with the low-orbit satellite. All UAVs can communicate with neighbor UAVs under the condition of no interference, and the network controller can control all UAVs. There are multiple data flows in the network. After being interfered by spectrum, the transmission of the data flows is affected. The network controller can make a routing decision for the entire network based on global network information, and plan the transmission path selection and network bandwidth allocation of the data flows. Figure 1 As shown in FIG. 1, the specific implementation steps of the routing decision are as follows:

[0046] 1. The network controller interacts with the UAVs to obtain global network information, including UAV positions, interference conditions, and data flows to be transmitted in the network.

[0047] 2. Based on the global network information, the network controller formulates an initial routing scheme for all data flows, including calculating the transmission path and allocating the network bandwidth.

[0048] 2.1 For all data flows, the Dijkstra algorithm is used to calculate the shortest time delay path from the starting point to the ending point as the transmission path.

[0049] 2.2 Since there can be several data flows passing through the same node, and different data flows have different demands on network bandwidth, the network controller needs to allocate network bandwidth to data flows. For any two nodes u and v, define C(u, v) as the upper bound of network bandwidth between u and v. When there is only one data flow from node u to node v, if C(u, v) is greater than the demand of the data flow, the network bandwidth allocated to the data flow is equal to its demand, otherwise, C(u, v) is allocated to the data flow entirely. When there are multiple data flows from node u to node v, if C(u, v) is greater than the sum of the demands of these data flows, the network bandwidth allocated to each data flow is equal to its demand; otherwise, C(u, v) is allocated to these data flows proportionally according to their demands. After the network bandwidth allocation scheme for all nodes is calculated, the actual available network bandwidth for a data flow is equal to the minimum of the network bandwidth allocated to it by all nodes on its path.

[0050] 3. The network controller evaluates the performance of the initial routing scheme.

[0051] 3.1 For any data flow F in the network, i its delay performance S(h i ) is evaluated and represented as:

[0052]

[0053] where h i represents the end-to-end delay of data flow F i , D i represents the expected delay of F i , and exp represents the exponential function with the natural constant e as the base.

[0054] 3.2 For any data flow F in the network, i its network bandwidth performance S(f i ) is evaluated and represented as:

[0055]

[0056] where R i represents the network bandwidth demand of data flow F i .

[0057] 3.3 Since different data flows have different demands on delay and network bandwidth, for any data flow F in the network, i its routing performance S(f i ) is equal to the weighted sum of its delay performance S(h i ) and its network bandwidth performance S(f i ), and is represented as:

[0058] S i = λ i S(h i )+(1-λ i )S(f i )

[0059] where λ i represents the weight adjustment parameter of data flow F i .

[0060] 3.4 Evaluate the comprehensive routing performance S of the initial routing scheme based on the routing performance of each data flow, S is the sum of the routing performance of all data flows.

[0061] 4. The network controller iteratively optimizes the initial routing scheme using the simulated annealing framework to generate the final routing scheme.

[0062] 4.1 Set the initial temperature, termination temperature, cooling coefficient, maximum number of iterations of simulated annealing, and set the current temperature equal to the initial temperature.

[0063] 4.2 Based on the current routing scheme, generate a new routing scheme according to the following steps:

[0064] 4.2.1 Randomly select a data flow F m , whose transmission path is p m , the starting point and the ending point are s m and t m respectively. Construct an intermediate node set containing all nodes that meet the following conditions: the node and s m are connected; the node and t m are connected; the node is not on p m .

[0065] 4.2.2 Randomly select a node n from the intermediate node set, use Dijkstra algorithm to calculate the shortest delay path p m from s n1 to node n and the shortest delay path p m from node n to t n2 , combine p n1 and p n2 to get a new path

[0066] 4.2.3 Reallocate network bandwidth to all data flows according to the network bandwidth allocation scheme in step 2.2 to get a new routing scheme.

[0067] 4.3 According to the performance evaluation method of step 3, the performance of the new routing scheme is evaluated, and the Metropolis acceptance criterion is used to judge whether to accept the new routing scheme. If the comprehensive routing performance of the new routing scheme is greater than the comprehensive routing performance of the current routing scheme, the new routing scheme is accepted, otherwise, with a probability p accept The new routing scheme is accepted:

[0068]

[0069] Wherein, exp represents the exponential function with the natural constant e as the base, S new represents the comprehensive routing performance of the new routing scheme, S temp represents the comprehensive routing performance of the current routing scheme, and T represents the current temperature of the simulated annealing system.

[0070] 4.4 If the number of cycles at the current temperature reaches the maximum number of cycles, multiply the current temperature by the cooling coefficient to update the value of the current temperature, and jump to step 4.5, otherwise return to step 4.2.

[0071] 4.5 If the current temperature is less than the termination temperature, end the loop, and output the current routing scheme as the optimal routing scheme, otherwise return to step 4.2.

[0072] The effects of the present application can be further illustrated by the following simulation experiment.

[0073] The present application uses PyCharm 2022.1 software for simulation, and part of the simulation parameters are shown in Table 1. The center of the ground interference source is taken as the center, and the size of the simulation area is 800m*800m. A specified number of unmanned aerial vehicles are generated in the area, and the number of unmanned aerial vehicles is 30, 40, 50, 60 and 70 respectively, and the coordinate positions of the unmanned aerial vehicles are randomly generated. Then a specified number of data streams are generated, and the number of data streams is 50% and 100% of the number of unmanned aerial vehicles. The number of data streams is defined as 50% of the number of unmanned aerial vehicles in the low-load scenario, and the number of data streams is defined as 100% of the number of unmanned aerial vehicles in the high-load scenario, thereby forming 10 test scenarios with different numbers of unmanned aerial vehicles and data streams.

[0074] Table 1

[0075] Simulation parameters Set values Height of UAV 100m Channel gain per unit distance -60dB Transmit power of UAV 6W Transmit power of interference source 20W Ambient noise power -169dBm / Hz Channel bandwidth 5MHz Upper limit of transmission rate of UAV and low-orbit satellite 50Mbps

[0076] The present application uses average delay performance, average network bandwidth performance and comprehensive routing performance as performance indicators, wherein the average delay performance is the average value of the delay performance of all data streams, and the average network bandwidth performance is the average value of the network bandwidth performance of all data streams.

[0077] The experimental results of the first experiment and the second experiment are shown in the following table. Figures 2 to 7 The present application selects the precise algorithm Gurobi as one of the comparison schemes, which has the advantage of being able to calculate the precise solution, but the calculation time increases exponentially with the expansion of the network scale, which does not have practical application value and can only be used as a reference. The calculation time of the algorithm SA-ROP proposed by the present application is often within a few seconds. The experimental results show that the present application can quickly obtain an approximate solution in any case, especially in improving the network bandwidth performance, which is very suitable for large-scale unmanned aerial vehicle networks.

[0078] The above is only a case of the specific embodiment of the present application, and the protection scope of the present application is not limited to the above examples. All technical solutions meeting the idea of the present application belong to the protection scope of the present application. It should be pointed out that, for the engineering technicians in the field, some improvements and optimizations without departing from the principle of the present application should be regarded as the protection scope of the present application.

Claims

1. A method for routing in a UAV network under spectrum jamming based on low earth orbit satellite coordination, characterized in that: The unmanned aerial vehicle network comprises a network controller, a low-orbit satellite, a plurality of master unmanned aerial vehicles and a plurality of slave unmanned aerial vehicles, each master unmanned aerial vehicle is provided with a satellite communication module and can perform bidirectional communication with the low-orbit satellite, all unmanned aerial vehicles can communicate with neighbor unmanned aerial vehicles in the absence of interference, the network controller can control all unmanned aerial vehicles and, after being interfered, make a decision on the routing of the entire network based on global network information and make a transmission scheme for all data streams in the network, and the method specifically comprises the following steps: Step (1) The network controller acquires global network information: the network controller interacts with the unmanned aerial vehicles to acquire information such as the positions of the unmanned aerial vehicles, interference conditions and data streams to be transmitted in the network; Step (2) An initial routing scheme is made: based on the global network information, the network controller first calculates transmission paths for all data streams and then allocates network bandwidth to obtain an initial routing scheme for the data streams; Step (3) The performance of the initial routing scheme is evaluated: the network controller respectively evaluates the delay performance and network bandwidth performance of all data streams, calculates the routing performance of the data streams in combination with a weight adjustment parameter and further evaluates the comprehensive routing performance of the initial routing scheme; Step (4) The routing scheme is optimized: based on a simulated annealing framework, the network controller iteratively optimizes the path selection and network bandwidth allocation of the initial routing scheme until a termination condition is met to obtain a final routing scheme, and the optimization specifically comprises: Step (4.1) An initial temperature, a termination temperature, a cooling coefficient, a maximum number of cycles and a current temperature equal to the initial temperature are set; Step (4.2) A new routing scheme is generated based on the current routing scheme according to the following steps: Step (4.2.1) randomly select a data flow F m with transmission path p m , start point s m and end point t m , construct an intermediate node set containing all nodes that satisfy the following conditions: the node is connected with s m ; the node is connected with t m ; the node is not on p m ; Step (4.2.2) randomly select a node n in the intermediate node set, use Dijkstra algorithm to calculate the shortest delay path p from s to node n m and the shortest delay path p from node n to t m , combine p n1 and p n2 to get a new path p n1 . n2 ​ Step (4.2.3) Network bandwidth is re-allocated to all data streams according to the network bandwidth allocation scheme in step (2) to obtain a new routing scheme; Step (4.3) evaluates the performance of the new routing scheme according to the performance evaluation method of step (3), and uses the Metropolis acceptance criterion to determine whether to accept the new routing scheme. If the comprehensive routing performance of the new routing scheme is greater than the comprehensive routing performance of the current routing scheme, the new routing scheme is accepted; otherwise, the new routing scheme is accepted with a probability p accept The new routing scheme is accepted: where exp denotes the exponential function with base of the natural constant e, S new represents the comprehensive routing performance of the new routing scheme, S temp represents the comprehensive routing performance of the current routing scheme, T represents the current temperature of the simulated annealing system; Step (4.4) If the number of cycles at the current temperature reaches the maximum number of cycles, the current temperature is multiplied by the cooling coefficient to update the value of the current temperature, and the process is jumped to step (4.5), otherwise, the process is returned to step (4.2); Step (4.5) If the current temperature is less than the termination temperature, the cycle is ended, and the current routing scheme is output as the optimal routing scheme, otherwise, the process is returned to step (4.2).

2. The method of claim 1, wherein, In step (2), the detailed steps of making the initial routing scheme are as follows: Step (2.1) For all data streams, a shortest delay path from the starting point to the ending point is calculated as a transmission path by using a Dijkstra algorithm. Step (2.2) Since several data streams can pass through the same node and different data streams have different network bandwidth requirements, the network controller needs to allocate network bandwidth to the data streams. For any two nodes u and v, define C(u, v) as the upper limit of network bandwidth between u and v. When there is only one data stream from node u to node v, if C(u, v) is greater than the requirement of the data stream, the network bandwidth allocated to the data stream is equal to its requirement, otherwise, C(u, v) is allocated to the data stream entirely. When there are multiple data streams from node u to node v, if C(u, v) is greater than the sum of network bandwidth requirements of these data streams, the network bandwidth allocated to each data stream is equal to its network bandwidth requirement; otherwise, C(u, v) is allocated to these data streams in proportion to their network bandwidth requirements. After calculating the network bandwidth allocation scheme of all nodes, the actual available network bandwidth of a data stream is equal to the minimum value of the network bandwidth allocated to it by all nodes on its path. 3.The method of claim 1, wherein, The detailed steps of step (3) for evaluating the performance of the initial routing scheme are as follows: Step (3.1) is performed for any data flow F in the network i , whose delay performance S(h i ) is evaluated and expressed as: where h i Represents data flow F i The end-to-end delay, D i F i The expected delay of , exp represents the exponential function with the natural constant e as the base; Step (3.2) is performed for any data flow F in the network i , whose network bandwidth performance S(f i ) is evaluated and expressed as: wherein R i represents the network bandwidth requirement of the data stream F i . Step (3.3) Since the requirements of different data streams for latency and network bandwidth are not the same, for any data stream F i its routing performance S(f i ) is equal to the weighted sum of the latency performance S(h i ) and the network bandwidth performance S(f i ), expressed as: S i = λ i S(hi)+(1-λ i )S(f i ) where λ i represents a weight adjustment parameter for the data stream F i . Step (3.4) Based on the routing performance of each data stream, the comprehensive routing performance S of the initial routing scheme is evaluated, and S is the sum of the routing performance of all data streams.

Images