A method for unmanned aerial vehicle cooperative service based on intelligent lamp pole network

Abstract

The application provides a kind of unmanned aerial vehicle cooperative service method based on wisdom lamp pole network, belongs to intelligent transportation and internet of things technical field.The method is formed by deploying wisdom lamp pole node to form distributed network, each node is equipped with environmental perception, communication, edge computing module and network interface.After unmanned aerial vehicle access, it is registered to gateway node and reports task planning and resource demand.Wisdom lamp pole network summarizes information, and dynamically generates cooperative service strategy for unmanned aerial vehicle through distributed cooperative calculation.In task execution phase, adaptive communication relay and calculation offload service are provided according to the strategy, and dynamically updated.The method also contains emergency cooperative response mechanism, through event response contract and PBFT consensus algorithm, unmanned aerial vehicle adjusts task strategy autonomously, and cooperates to complete event monitoring and data acquisition.The application can improve the stability of unmanned aerial vehicle communication, computing capacity and task execution efficiency, and expand its application range.

Description

Technical Field

[0001] This invention relates to the fields of intelligent transportation and the Internet of Things, and in particular to a method for drone collaborative services based on a smart light pole network. Background Technology

[0002] With the rapid development of drone technology, drones have been widely used in many fields such as logistics delivery, environmental monitoring, disaster relief, and traffic management. However, drones face many challenges during operation, such as unstable communication links, limited computing resources, and insufficient endurance. These problems limit the performance and application scope of drones.

[0003] Meanwhile, smart light poles, as an emerging urban infrastructure, are gradually becoming more widespread in cities. Smart light poles not only possess traditional lighting functions but also integrate various intelligent devices, such as environmental sensing modules, communication modules, and edge computing modules, enabling real-time environmental perception, data transmission, and local computing processing. The distributed architecture of smart light pole networks gives them unique advantages in terms of coverage and resource distribution, providing new ideas and possibilities for solving many problems faced by drones.

[0004] Currently, although some research has explored collaborative working modes between UAVs and ground infrastructure, most of these studies focus on single-function collaboration, such as communication relay or simple data transmission, lacking comprehensive consideration of the multifaceted needs of UAVs and systematic solutions. Furthermore, existing solutions often struggle to achieve flexible and efficient resource scheduling and service optimization when facing dynamically changing mission requirements and network environments for UAVs. Summary of the Invention

[0005] The purpose of this invention is to provide a drone collaborative service method based on a smart light pole network. By leveraging the distributed sensing, communication, and edge computing capabilities of the smart light pole network, it provides dynamic collaborative services for drones, including communication relay and computation offloading, thereby improving their performance and application value in complex environments. At the same time, it enables rapid collaborative response to emergencies and expands the application scope of drones.

[0006] To achieve the above objectives, the present invention provides a drone collaborative service method based on a smart light pole network, comprising the following steps: Step S1: Deploy multiple smart light pole nodes to form a distributed sensing and communication network. This network is a smart light pole network. Each smart light pole node is equipped with an environmental sensing module, a communication module, an edge computing module, and a network interface. Each smart light pole node collects environmental parameters through the environmental sensing module and obtains network topology and node load information through the communication module. Step S2: The drone connected to the smart light pole network initiates a registration request to the nearest smart light pole gateway node. During the registration process, it reports its flight mission plan and resource requirements. The smart light pole gateway node synchronizes the drone's registration information, flight mission plan and resource requirements to the entire smart light pole network. Step S3: The smart light pole network summarizes the environmental parameters, network topology, and node load information collected by each smart light pole node in step S1, and combines them with the task planning and resource requirements reported by the UAV in step S2. Through distributed collaborative computing, a collaborative service strategy is dynamically generated for the UAV. Step S4, during the drone's mission execution phase, the smart light pole network provides adaptive communication relay and computing offloading services for the drone through real-time data interaction and resource scheduling between smart light pole nodes, based on the collaborative service strategy generated in step S3. The communication relay dynamically adjusts the relay smart light pole nodes based on the link maintenance strategy, and the computing offloading service allocates edge computing resources based on the task allocation strategy.

[0007] Preferably, in step S1, the environmental parameters include wind speed. and precipitation Network topology information includes the connection relationships between smart light pole nodes, and node load information includes CPU utilization. and bandwidth utilization .

[0008] Preferably, in step S2, flight mission planning includes route and time window information, and resource requirements include communication bandwidth requirements and computing power requirements.

[0009] Preferably, in step S3, the collaborative service strategy includes a communication link maintenance strategy and a computing task allocation strategy.

[0010] Preferably, in step S3, the generation of the communication link maintenance strategy specifically includes: Predicting drones in the future The communication link status within the time frame is determined using a link quality predictor based on a long short-term memory network. The link quality predictor predicts the status at time [time value missing]. The input feature vector is Its elemental composition is as follows: ; in, Indicates time The relative position vectors of the drone and all smart light pole nodes. Indicates time The signal-to-noise ratio vector of the communication link between the drone and all smart light pole nodes. Indicates time Network delay matrix of each link in the smart light pole network Indicates time Wind speed collected by the environmental sensing module and precipitation , Indicates time The flight velocity vector of the drone Indicates time Bandwidth utilization of each smart light pole node; The link quality predictor output of a Long Short-Term Memory (LSTM) network will predict the future. Link quality prediction at each time step If any link quality prediction value Below the preset threshold This triggers a link switching decision, allowing the drone to pre-select a link switcher. A new communication path formed by smart light pole nodes , Indicates the node number of the smart light pole; The path selection must meet the communication bandwidth requirements of the UAV mission, that is, the sum of the available bandwidth of all nodes in the path must meet the following requirements: ; in, Indicates the first in the path Smart light pole nodes Total bandwidth capacity Indicates the first in the path Smart light pole nodes bandwidth utilization Indicates drone The communication bandwidth requirements.

[0011] Preferably, a new communication path is selected. The decision modeling is a reinforcement learning problem, with the objective of maximizing the value function, satisfying: ; in, Indicates time The status includes the smart light pole network topology, link load, and drone location. Indicates time The action, namely the number of the next hop relay smart light pole node, Indicates time Execute action The immediate reward afterwards Indicates the discount factor. Operator for mathematical expectation; The calculation formula is as follows: ; in, Indicates selection of smart light pole node. Expected throughput Indicates selection of smart light pole node. Expected communication latency, Indicates the connection from the previous smart light pole node. Switch to The cost, , , These represent throughput weight, latency weight, and switching cost weight, respectively.

[0012] Preferably, in step S3, the generation of the task allocation strategy specifically includes: drones The computational task is modeled as a tuple. ,in, This indicates the total number of CPU cycles required to complete the task. This indicates the amount of input data for the task. This indicates the maximum allowable completion time for the task; The smart street light network transforms the task allocation problem into an optimization problem aimed at minimizing the total task completion time, with the objective function being: ; The constraints include: ; ; ; in, This represents the set of all smart light pole nodes with available edge computing capabilities. Indicates drone With smart light pole nodes Communication bandwidth between them This indicates that the node is assigned to the smart light pole. A subset of tasks Indicates smart light pole nodes Available CPU computing power Indicates smart light pole nodes CPU utilization Indicates the length of a unit of time. Indicates drone The reported communication bandwidth requirements.

[0013] Preferably, for the computational task allocation optimization problem, combining the edge computing characteristics of smart street light networks and the dynamic nature of drone tasks, an online algorithm based on Lyapunov optimization is used for real-time solution; the online algorithm based on Lyapunov optimization at each time step... Minimize the upper bound of the Lyapunov drift plus penalty function, satisfying: ; in, Indicates time The queue backlog of computing tasks for each UAV at that time. Corresponding drones Task queue length, This indicates Lyapunov drift. It is a second-order Lyapunov function. Indicates control parameters, Indicates drone The task is at all times The estimated execution time; .

[0014] Therefore, this invention employs the aforementioned UAV collaborative service method based on a smart light pole network. First, the environmental perception module of the smart light pole node collects environmental parameters, and the communication module obtains network topology and node load information, achieving comprehensive perception of the network status and environment, providing the UAV with more accurate communication and computing resource allocation. Second, the UAV registers with the smart light pole gateway node and reports its task plan and resource requirements, enabling the entire smart light pole network to dynamically generate collaborative service strategies. This distributed collaborative computing approach improves the accuracy and flexibility of decision-making compared to traditional single-node decision-making. During the UAV's task execution phase, the smart light pole network, based on the collaborative service strategy, provides adaptive communication relay and computing offloading services to the UAV through real-time data interaction and resource scheduling between nodes. The service strategy is dynamically updated according to changes in the UAV's location and network status, effectively improving the UAV's communication stability and computing power in complex environments, and enhancing the reliability and efficiency of task execution. Furthermore, this invention also possesses collaborative response capabilities to emergencies, further expanding the application scenarios and value of UAVs. Attached Figure Description

[0015] Figure 1 This is a flowchart of a drone collaborative service method based on a smart light pole network according to the present invention; Figure 2 Generate a flowchart for the communication link maintenance strategy; Figure 3 Generate a flowchart for the strategy of assigning computational tasks. Detailed Implementation

[0016] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0017] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0018] Example 1 like Figure 1 As shown, a drone collaborative service method based on a smart light pole network includes the following steps: Step S1: Deploy multiple smart light pole nodes to form a distributed sensing and communication network. This network is a smart light pole network. Each smart light pole node is equipped with an environmental sensing module, a communication module, an edge computing module, and a network interface. Each smart light pole node collects environmental parameters through the environmental sensing module and obtains network topology and node load information through the communication module.

[0019] Environmental parameters include wind speed and precipitation Network topology information includes the connection relationships between smart light pole nodes, and node load information includes CPU utilization. and bandwidth utilization .

[0020] In step S2, the drone connected to the smart light pole network initiates a registration request to the nearest smart light pole gateway node. During the registration process, it reports its flight mission plan and resource requirements. The flight mission plan includes route and time window information, and the resource requirements include communication bandwidth requirements and computing power requirements. The smart light pole gateway node synchronizes the drone's registration information, flight mission plan, and resource requirements to the entire smart light pole network.

[0021] Step S3: The smart light pole network summarizes the environmental parameters, network topology, and node load information collected by each smart light pole node in Step S1, and combines them with the task planning and resource requirements reported by the UAV in Step S2. Through distributed collaborative computing, a collaborative service strategy is dynamically generated for the UAV.

[0022] The collaborative service strategy includes communication link maintenance strategy and computing task allocation strategy.

[0023] like Figure 2 As shown, the generation of the communication link maintenance strategy specifically includes: Predicting drones in the future The communication link status within the time frame is determined using a link quality predictor based on a long short-term memory network. The link quality predictor predicts the status at time [time value missing]. The input feature vector is Its elemental composition is as follows: ; in, Indicates time The relative position vectors of the drone and all smart light pole nodes are generated by obtaining distance measurements through ultra-wideband two-way ranging technology between the drone and at least three smart light pole nodes at different locations, and then calculating the three-dimensional coordinates of the drone using triangulation (this is existing technology and will not be described in detail here). Indicates time Signal-to-noise ratio vector of the communication link between the drone and all smart light pole nodes; Indicates time The network delay matrix of each link in the smart light pole network is related to the node connection relationship and node load in the network topology (the higher the load, the greater the delay). Indicates time Wind speed collected by the environmental sensing module and precipitation ; Indicates time The drone's flight velocity vector, including the magnitude and direction of the velocity; Indicates time Bandwidth utilization of each smart light pole node.

[0024] The link quality predictor output of a Long Short-Term Memory (LSTM) network will predict the future. Link quality prediction at each time step If any link quality prediction value Below the preset threshold This triggers a link switching decision, allowing the drone to pre-select a link switcher. A new communication path formed by smart light pole nodes , This indicates the node number of the smart light pole.

[0025] The path selection must meet the communication bandwidth requirements of the UAV mission, that is, the sum of the available bandwidth of all nodes in the path must meet the following requirements: ; in, Indicates the first in the path Smart light pole nodes Total bandwidth capacity Indicates the first in the path Smart light pole nodes bandwidth utilization Indicates drone The communication bandwidth requirements.

[0026] Select a new communication path The decision modeling is a reinforcement learning problem, with the objective of maximizing the value function, satisfying: ; in, Indicates time The status includes the smart light pole network topology (node ​​connection relationships), link load, and drone location. Indicates time The action, namely the number of the next hop relay smart light pole node, Indicates time Execute action The immediate reward afterwards Indicates the discount factor. Operator for mathematical expectation; The calculation formula is as follows: ; in, Indicates selection of smart light pole node. Expected throughput Indicates selection of smart light pole node. Expected communication latency, Indicates the connection from the previous smart light pole node. Switch to The cost, , , These represent throughput weight, latency weight, and switching cost weight, respectively. ; ; ; in, This indicates the available bandwidth of the current candidate smart light pole relay node. This indicates the communication bandwidth requirements of the drone. This represents the basic weighting coefficient (preset range 0.3-0.5). This represents the measured communication latency of the current candidate smart light pole relay node. Indicates the remaining time for the task. This represents the basic weighting coefficient (preset range 0.3-0.5). Indicates the current flight speed of the drone. Indicates the maximum design speed of the drone. Indicates recent Number of switching times within a time period This indicates the maximum allowed number of handovers (preset based on network stability requirements). This represents the basic weighting coefficient (preset range 0.2-0.4).

[0027] Ultimately used in the reward function , , Normalization is required (to ensure) ), and every interval Recalculate the weights.

[0028] like Figure 3 As shown, the generation of the computational task allocation strategy specifically includes: drones The computational task is modeled as a tuple. ,in, This represents the total number of CPU cycles required to complete the task, satisfying... , Indicates drone The reported computing power requirements Indicates drone The deadline for the task. This indicates the amount of input data for the task. This indicates the maximum allowable completion time for the task; The smart street light network transforms the task allocation problem into an optimization problem aimed at minimizing the total task completion time, with the objective function being: ; The constraints include: ; ; ; in, This represents the set of all smart light pole nodes with available edge computing capabilities (the elements in the set are smart light pole node numbers, and light poles with computing resources that can participate in task offloading are selected). Indicates drone With smart light pole nodes The communication bandwidth between nodes (its value is related to the connection relationship between nodes in the network topology and is affected by the performance of the wireless channel and communication module). This indicates that the node is assigned to the smart light pole. A subset of tasks (the elements in the set are drone task numbers, representing the list of tasks that the light pole needs to handle). Indicates smart light pole nodes Available CPU computing power Indicates smart light pole nodes CPU utilization This indicates the length of a unit of time (the time granularity of task allocation and scheduling). Indicates drone Reported communication bandwidth requirements (characterizing the expected acquisition rate of light pole communication resources when the UAV transmits mission data).

[0029] The calculation formula is as follows: ; in, Indicates smart light pole nodes The total computing power is the upper limit of computing resources for the light pole hardware design.

[0030] To address the computational task allocation optimization problem, this paper combines the edge computing characteristics of smart street light networks with the dynamic nature of drone missions, employing an online algorithm based on Lyapunov optimization for real-time solution to avoid reliance on future information. The online algorithm based on Lyapunov optimization performs real-time calculations at each time step. Minimize the upper bound of the Lyapunov drift plus penalty function, satisfying: ; in, Indicates time The queue backlog of computing tasks for each UAV at that time, among which Corresponding drones Task queue length, This indicates Lyapunov drift. This is a quadratic Lyapunov function, used to characterize the stability of the task queue in a smart street light network. This represents a control parameter (positive number) used to balance the latency requirements of drone missions with the computational resource consumption of smart light pole nodes. Indicates drone The task is at all times The estimated execution time; .

[0031] The algorithm described above dynamically adjusts the task allocation scheme by sensing changes in node load in the smart light pole network and changes in communication links caused by the movement of drones in real time, so that the optimization process is adapted to the distributed architecture of the smart light pole network and the mobility characteristics of drones.

[0032] Step S4, during the drone's mission execution phase, the smart light pole network, based on the collaborative service strategy generated in step S3, provides adaptive communication relay and computation offloading services for the drone through real-time data interaction and resource scheduling between smart light pole nodes. The communication relay dynamically adjusts the relay smart light pole nodes (i.e., the smart light pole nodes that undertake relay functions) based on the link maintenance strategy. The computation offloading service allocates edge computing resources based on the task allocation strategy and dynamically updates the service strategy as the drone's location and network status change.

[0033] The above methods also include coordinated response to emergencies: when any smart light pole node... When an emergency (including a traffic accident) is detected, an event response contract is generated: ; The Practical Byzantine Fault Tolerance (PBFT) algorithm is used as the consensus algorithm for broadcasting within the smart light pole network. The broadcasting achieves data transmission through the communication module between smart light pole nodes, ensuring that the contract is received by all nodes in the smart light pole network.

[0034] The consensus algorithm execution process is as follows: smart light pole nodes As the master node initiates a proposal, when more than [number] proposals are received... After the same confirmation from each smart light pole node, the event response contract... This is the consensus reached by the smart light pole network.

[0035] In the event response contract: The event type (e.g., traffic accident, fire); Location of the event (latitude and longitude coordinates); This represents the event priority (values ​​range from 1 to 5, with higher values ​​indicating higher priority). Specify the type of data to be collected (e.g., high-resolution images, temperature data).

[0036] Upon receiving the contract, the drone node adjusts its flight path according to the location of the event in the contract, reorders the task execution order according to the event priority, and configures the sensor acquisition parameters according to the data requirements. In this way, it can autonomously adjust its task strategy and collaboratively complete the monitoring and data collection of the event.

[0037] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0038] Therefore, the present invention adopts the above-mentioned drone collaborative service method based on smart light pole network. By leveraging the distributed sensing, communication and edge computing capabilities of smart light pole network, it provides dynamic collaborative services for drones, including communication relay and computing offloading, thereby improving their performance and application value in complex environments. At the same time, it enables rapid collaborative response to emergencies and expands the application scope of drones.

[0039] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for drone collaborative services based on a smart light pole network, characterized in that, Includes the following steps: Step S1: Deploy multiple smart light pole nodes to form a distributed sensing and communication network. This network is a smart light pole network. Each smart light pole node is equipped with an environmental sensing module, a communication module, an edge computing module, and a network interface. Each smart light pole node collects environmental parameters through the environmental sensing module and obtains network topology and node load information through the communication module. Step S2: The drone connected to the smart light pole network initiates a registration request to the nearest smart light pole gateway node. During the registration process, it reports its flight mission plan and resource requirements. The smart light pole gateway node synchronizes the drone's registration information, flight mission plan and resource requirements to the entire smart light pole network. Step S3: The smart light pole network summarizes the environmental parameters, network topology, and node load information collected by each smart light pole node in step S1, and combines them with the task planning and resource requirements reported by the UAV in step S2. Through distributed collaborative computing, a collaborative service strategy is dynamically generated for the UAV. Collaborative service strategies include communication link maintenance strategies and computing task allocation strategies; The generation of the communication link maintenance strategy includes: predicting the communication link status of the UAV in the future time period, using a link quality predictor based on a long short-term memory network for prediction, and if the predicted link quality is lower than a preset threshold, triggering a link switching decision to pre-select a new communication path that meets the UAV's communication bandwidth requirements. The generation of the computational task allocation strategy includes: modeling the UAV's computational tasks as tuples containing the total number of CPU cycles, the amount of input data, and the maximum completion latency; and constructing the task allocation problem as an optimization problem with the objective of minimizing the total task completion time, which is solved by allocating a subset of tasks to each smart light pole node. Step S4, during the drone's mission execution phase, the smart light pole network provides adaptive communication relay and computing offloading services for the drone through real-time data interaction and resource scheduling between smart light pole nodes, based on the collaborative service strategy generated in step S3. The communication relay dynamically adjusts the relay smart light pole nodes based on the communication link maintenance strategy, and the computing offloading service allocates edge computing resources based on the task allocation strategy.

2. The method for drone collaborative services based on a smart light pole network according to claim 1, characterized in that, In step S1, environmental parameters include wind speed. and precipitation Network topology information includes the connection relationships between smart light pole nodes, and node load information includes CPU utilization. and bandwidth utilization .

3. The method for drone collaborative services based on a smart light pole network according to claim 1, characterized in that, In step S2, flight mission planning includes route and time window information, and resource requirements include communication bandwidth requirements and computing power requirements.

4. The method for drone collaborative services based on a smart light pole network according to claim 1, characterized in that, In step S3, the generation of the communication link maintenance strategy specifically includes: Predicting drones in the future The communication link status within the time frame is determined using a link quality predictor based on a long short-term memory network. The link quality predictor predicts the status at time [time value missing]. The input feature vector is Its elemental composition is as follows: ； in, Indicates time The relative position vectors of the drone and all smart light pole nodes. Indicates time The signal-to-noise ratio vector of the communication link between the drone and all smart light pole nodes. Indicates time Network delay matrix of each link in the smart light pole network Indicates time Wind speed collected by the environmental sensing module and precipitation , Indicates time The flight velocity vector of the drone Indicates time Bandwidth utilization of each smart light pole node; The link quality predictor output of a Long Short-Term Memory (LSTM) network will predict the future. Link quality prediction at each time step If any link quality prediction value Below the preset threshold This triggers a link switching decision, allowing the drone to pre-select a link switcher. A new communication path formed by smart light pole nodes , Indicates the node number of the smart light pole; The path selection must meet the communication bandwidth requirements of the UAV mission, that is, the sum of the available bandwidth of all nodes in the path must meet the following requirements: ； in, Indicates the first in the path Smart light pole nodes Total bandwidth capacity Indicates the first in the path Smart light pole nodes bandwidth utilization Indicates drone The communication bandwidth requirements.

5. The method for drone collaborative services based on a smart light pole network according to claim 4, characterized in that, Select a new communication path The decision modeling is a reinforcement learning problem, with the objective of maximizing the value function, satisfying: ； in, Indicates time The status includes the smart light pole network topology, link load, and drone location. Indicates time The action, namely the number of the next hop relay smart light pole node, Indicates time Execute action The immediate reward afterwards Indicates the discount factor. Operator for mathematical expectation; The calculation formula is as follows: ； in, Indicates selection of smart light pole node. Expected throughput Indicates selection of smart light pole node. Expected communication latency, Indicates the connection from the previous smart light pole node. Switch to The cost, , , These represent throughput weight, latency weight, and switching cost weight, respectively.

6. The method for drone collaborative services based on a smart light pole network according to claim 5, characterized in that, In step S3, the generation of the task allocation strategy specifically includes: drones The computational task is modeled as a tuple. ,in, This indicates the total number of CPU cycles required to complete the task. This indicates the amount of input data for the task. This indicates the maximum allowable completion time for the task; The smart street light network transforms the task allocation problem into an optimization problem aimed at minimizing the total task completion time, with the objective function being: ； The constraints include: ； ； ； in, This represents the set of all smart light pole nodes with available edge computing capabilities. Indicates drone With smart light pole nodes Communication bandwidth between them This indicates that the node is assigned to the smart light pole. A subset of tasks Indicates smart light pole nodes Available CPU computing power Indicates smart light pole nodes CPU utilization Indicates the length of a unit of time. Indicates drone The reported communication bandwidth requirements.

7. The method for drone collaborative services based on a smart light pole network according to claim 6, characterized in that, To address the computational task allocation optimization problem, and considering the edge computing characteristics of smart street light networks and the dynamic nature of drone tasks, an online algorithm based on Lyapunov optimization is employed for real-time solution. The online algorithm based on Lyapunov optimization performs real-time calculations at each time step. Minimize the upper bound of the Lyapunov drift plus penalty function, satisfying: ； in, Indicates time The queue backlog of computing tasks for each UAV at that time. Corresponding drones Task queue length, This indicates Lyapunov drift. It is a second-order Lyapunov function. Indicates control parameters, Indicates drone The task is at all times The estimated execution time; 。

Images