CN-121985406-A - Unmanned plane network-oriented self-adaptive time synchronization method

CN121985406ACN 121985406 ACN121985406 ACN 121985406ACN-121985406-A

Abstract

The invention discloses a self-adaptive time synchronization method for an unmanned plane network, which relates to the technical field of unmanned plane self-organizing network time synchronization and comprises the steps of S1, initializing, S2, node network access and neighbor discovery, S3, master node selection, S4, centralized time synchronization, S5, TDMA time slot allocation and frame control, S6, distributed synchronization, and S7, synchronization quality evaluation and rollback switching, wherein the master node is selected by calculating the master node scoring weight according to a neighbor information table, the slave node performs time synchronization according to the broadcasting of the master node, the slave node performs time synchronization according to the broadcasting of the slave node, the S7, synchronization quality evaluation and rollback switching are performed when the synchronization rollback condition is met, and the master node is selected by S3 when the master node reselection condition is met and the current master node does not participate in the master node selection. The invention adopts the self-adaptive switching of the TDMA mechanism and the centralized synchronization and the distributed synchronization, reduces the collision probability and the energy consumption of the nodes, and avoids the error accumulation of a single synchronization mode.

Inventors

CHEN LINTAO
CHEN YINCHAO
ZHANG CHENGZHE
LIN ZUMING
ZHANG YI

Assignees

厦门大学

Dates

Publication Date: 20260505
Application Date: 20260206
Priority Date: 20260123

Claims (10)

1.An unmanned aerial vehicle network-oriented self-adaptive time synchronization method is characterized by comprising the following steps: s1, initializing network node parameters of an unmanned aerial vehicle; s2, node networking and neighbor discovery, namely, each unmanned plane node sends a HELLO data packet, and each node updates neighbor information of a neighbor information table by using the received HELLO data packet; S3, the master node selecting step, wherein each node calculates the master node grading weight according to the neighbor information table and broadcasts grading data packets comprising the master node grading weight, each node updates the master node grading weight of the neighbor information table by using the received grading data packets, and takes the node with the highest grading weight of the master node in the neighbor information table as the master node candidate; s4, a centralized time synchronization step, wherein a main node broadcasts a reference data packet comprising a time stamp, and a sub-node receives the reference data packet and performs time synchronization with the main node; s5, a TDMA time slot allocation and frame control step, wherein a main node broadcasts a frame data packet in a time slot of the frame head of the TDMA, and each sub-node receives the frame data packet, updates local frame parameters and performs frame calibration; S6, a distributed synchronization step, wherein the sub-nodes broadcast distributed synchronization data packets comprising a consistency logic clock in the allocated time slots; S7, a step of synchronous quality evaluation and rollback switching, namely, the master node receives the distributed synchronous data packet of the child node and calculates a synchronous residual error quantifying clock deviation between the child node and the master node, when a synchronous rollback condition based on the synchronous residual error is met, S4 is carried out, and when a master node reselection condition is met, S3 is carried out, and the current master node does not participate in master node selection.
2. The unmanned network-oriented adaptive time synchronization method of claim 1, wherein the master node scoring weights are expressed as: ; ; Wherein, the Representing nodes Is a master node scoring weight; Representing nodes For measuring the number of node neighbors; Representing nodes Based on the remaining energy ratio; Representing nodes Is not limited by the movement stability of the device; Representing nodes Measuring the average distance between the node and the neighbor; Representing nodes Based on the packet reception rate; 、、、 And Representing the weight coefficient; Representing an upper limit cutoff; indicating the expiration time of the link and taking the value as the boundary condition of the link Least positive root or (2) , Representing nodes And node Is used for the relative position of the two, Representing nodes And node Is used for the control of the relative velocity of (c) in the system, Representing the communication radius of the unmanned aerial vehicle node, The time is represented by the time period of the day, Representing taking a model; Representing nodes Neighbor information table of (a); Representing taking the minimum value.
3. The unmanned aerial vehicle network-oriented adaptive time synchronization method according to claim 2, wherein the step S3 further comprises disambiguating by using deterministic decision rules, specifically: When the scoring weights of the main nodes of the two nodes are smaller than a preset scoring difference threshold, the nodes with higher scores are sequentially compared according to the priority order of connectivity, energy factors and link stability, and the nodes with higher scores are used as main node candidates.
4. The unmanned network-oriented adaptive time synchronization method of claim 1, wherein the child node receives the reference data packet and performs time synchronization with the master node, specifically as follows: Computing sub-nodes The time phase offset from the master node is expressed as: ; Wherein, the Representing a time phase deviation; Representing a unidirectional transmission delay compensation term; Representing child nodes Local timestamp at reception time of master TIMESYN _head packet received ; Representing the first broadcast by the master node Round local time stamping; Child node For continuous Averaging the time phase deviation of the secondary reference broadcast to obtain a final deviation; Updating the child node with the final offset, expressed as: ; ; Wherein, the Representing updated nodes A time offset compensation parameter of (a); Representing nodes A time offset compensation parameter of (a); Representing the final offset; representing an offset correction step size; Representing updated child nodes Output time of (2); Representing child nodes Output time of (2).
5. The unmanned network-oriented adaptive time synchronization method of claim 1, wherein the frame parameters of the time division multiple access TDMA are expressed as: Wherein, the method comprises the steps of, Representing a transmission frame period; representing the number of slots per frame; representing the slot length; Representing a guard interval; Representing an end-of-frame maintenance window, a child node advancing before the arrival of an assigned slot Time wake-up, delay after the end of a slot And closing the transmission, wherein the rest time enters a low-power-consumption dormant state or performs the rest information transmission task, and the rest nodes keep the receiving state in the time slot.
6. The adaptive time synchronization method for an unmanned aerial vehicle network according to claim 5, wherein the end-of-frame maintenance window is used for receiving broadcast signals of joining and exiting of a sub-node, the newly joined sub-node monitors a main node frame control data packet to synchronize a current frame state and time before joining the network, then a CSMA/CA mechanism is adopted to send a network joining request req_join in the end-of-frame maintenance window of the current frame, the main node updates a frame scheduling version and a frame time slot map after receiving the request, the sub-node ready to exit sends a network exiting request req_leave in the end-of-frame maintenance window, the main node marks the sub-node as an off-network state if the main node does not monitor a broadcast of a certain sub-node in a corresponding time slot in a continuous frame reaching an off-network threshold frame number, and updates a frame time slot table to release the time slot if the broadcast is not monitored within the off-network recovery threshold frame number.
7. The unmanned network-oriented adaptive time synchronization method of claim 1, wherein the distributed time synchronization step is specifically as follows: When the node At the receiving node When broadcasting distributed synchronous data packet, recording local hardware clock at receiving time and calculating local logic clock at receiving time, according to the local hardware clock at receiving time Updating its reception time logical clock value, expressed as: ; Wherein, the Representing updated nodes A local logic clock at a time of reception; Representing nodes A local logic clock at a time of reception; And Representing nodes respectively Sum node Confidence coefficient, initial value is 1, and corresponding node adds 1 after updating parameter successfully each time; Representing nodes Broadcast transmission to nodes A unidirectional transmission delay compensation term of (a); Representing nodes Is a transmission time logic clock; Node Updating the coherent logic clock, expressed as: ; ; ; Wherein, the Representing nodes Is a coherent logic clock of (a); representing updated nodes A clock frequency compensation parameter; Representing nodes A clock frequency compensation parameter; representing a local hardware clock; representing updated nodes A clock offset compensation parameter; Representing nodes A clock offset compensation parameter; represent the first Node at wheel end Updated logic clock values.
8. The unmanned network-oriented adaptive time synchronization method of claim 1, wherein the synchronization residual is expressed as: ; Wherein, the Representing nodes Synchronization residual with the master node; Representing nodes Broadcasting a unidirectional transmission delay compensation item transmitted to a main node; represent the first A wheel; Representing nodes A logic clock of a transmission timing of (a); A logic clock representing the time of receipt by the master node.
9. The unmanned aerial vehicle network-oriented self-adaptive time synchronization method of claim 1, wherein the synchronization rollback condition based on the synchronization residual is specifically that the distributed time synchronization operation is not less than the minimum resident frame number and any one of the following conditions is satisfied: ; ; ; ; Wherein, the Representing residual variances of all child nodes; representing a preset residual variance threshold; representing the maximum value in the synchronization residual errors of all the child nodes; Representing a preset maximum residual threshold; Representing the collision rate; Representing a preset collision rate threshold; Indicating the number of conflicting events, i.e. most recent The number of broadcast collision events from multiple nodes detected by a frame in one time slot; representing the number of slots per frame.
10. The unmanned aerial vehicle network-oriented adaptive time synchronization method of claim 1, wherein the main node reselection condition is specifically that the distributed time synchronization operation is not less than a minimum resident frame number and any one of the following conditions is satisfied: ; ; ; ; Wherein, the Representing the accumulated operation time of the master node; Representing a preset master node accumulated operation time threshold; Representing the energy of the master node; Representing a preset energy threshold; Representing the dynamic change degree of the network topology; representing a preset network topology dynamic change degree threshold; Indicating the latest The total number of network event joining and exiting in the frame; representing the transmission frame period.

Description

Unmanned plane network-oriented self-adaptive time synchronization method Technical Field The invention relates to unmanned aerial vehicle self-organizing network time synchronization technology, in particular to a self-adapting time synchronization method facing an unmanned aerial vehicle network. Background An unmanned aerial vehicle self-organizing Network (FANET) is used as a high-mobility and high-dynamic wireless self-organizing Network, the Network topology changes frequently along with flight and task formation, and the unmanned aerial vehicle self-organizing Network is often used for tasks such as reconnaissance monitoring, cooperative striking, emergency rescue, disaster monitoring and the like. Among these complex tasks, time synchronization is a crucial basis for the generation of synergistic effects by the unmanned aerial vehicle group. The time synchronization of the unmanned aerial vehicle self-organizing network is generally limited by the following steps that firstly, engineering resource constraints such as load, power consumption, cost and the like exist in tasks with specific requirements, requirements on a network related synchronization algorithm are easy to achieve, cost and energy consumption are low, secondly, unmanned aerial vehicle deployment is changed continuously along with the requirements of the tasks, certain differences exist in hardware circuits among nodes, the unmanned aerial vehicle deployment is affected by interference of an acting environment and a manufacturing process, certain time deviation is inevitably generated, the errors are accumulated along with multi-hop transmission, and thirdly, the unmanned aerial vehicle is generally in a motion state, and propagation condition changes, link delay fluctuation such as Doppler effect are caused by high-speed motion, so that synchronization error amplification is caused. The unmanned aerial vehicle network time synchronization is one of key factors of an unmanned aerial vehicle self-organizing network, and is a foundation for guaranteeing normal operation of the unmanned aerial vehicle network in various applications. Therefore, the requirements of high precision, high dynamic adaptability and low energy consumption are put forward for time synchronization among unmanned aerial vehicle nodes in the unmanned aerial vehicle network. Currently, time synchronization of unmanned aerial vehicle networks is mainly divided into two types, namely centralized synchronization, time service by satellite or time service by ground equipment with high time service precision or unmanned aerial vehicle reference node so as to reduce the influence of uncertainty of a transmitting end, and distributed synchronization. Traditional centralized time synchronization methods, such as reference broadcast synchronization (REFERENCE BROADCAST SYNCHRONIZATION, RBS), in which the RBS broadcasts data packets without local time stamps mainly through network reference nodes, and the receiving end compares multiple arrival receiving moments to achieve local time alignment, so that cross-region multi-hop synchronization can be achieved, but the accuracy can be gradually attenuated. The receiving end-receiving end synchronization mode eliminates the uncertainty of the transmitting end, but the uncertainty caused by message delay still has a great influence on the synchronization precision, and when the uncertainty of the message delay is larger, the generated time error is larger. The sensor network time synchronization protocol (Timing-sync Protocol for Sensor Networks, TPSN) uses a hierarchical topology to make a two-way handshake of a parent node and a child node, and synchronously aligns the clock of the whole network to a reference root node step by step. However, due to the strict dependence on the hierarchical structure, asymmetric time delay and progressive transmission under multiple hops can cause error accumulation, and additional overhead is brought to topology maintenance. The flooding time synchronization protocol (Flooding Time Synchronization Protocol, FTSP) selects the node with the smallest ID as the root node, and periodically broadcasts the synchronization data packet, and the receiving node uses linear regression to estimate clock drift to realize continuous correction, and broadcasts again to realize flooding synchronization after updating the local time. The bidirectional switching and the node on-network monitoring mode are used, so that the synchronization has higher precision and robustness, but in a high-dynamic network, the node on-off network can frequently trigger the node monitoring, and the communication overhead and the energy consumption are greatly increased. The delay measurement time synchronization algorithm (Delay Measurement Time Synchronization, DMTS) is consistent with the RBS to be a master-slave unidirectional time synchronization algorithm, the receiving node reduces transmission errors by estimating all del