CN-121619211-B - Off-network gateway dynamic election method and system based on edge AI and satellite ephemeris

Abstract

The invention discloses an off-network gateway dynamic election method and system based on an edge AI and a satellite ephemeris, which relate to the technical field of off-network communication networks and comprise the steps of forming a candidate gateway node set in an off-network ad hoc network, collecting multi-source running state information for each candidate gateway node in the candidate gateway node set, and constructing node state feature vectors according to a field arrangement sequence; based on the node state characteristic vector, intelligent reasoning is carried out on the edge side to generate gateway evaluation indexes, gateway availability constraint is applied to the gateway evaluation indexes by combining satellite ephemeris information to form constrained gateway evaluation indexes, dynamic off-network gateway election is carried out according to the constrained gateway evaluation indexes, off-network gateway nodes are determined, gateway role notification information is broadcast by the off-network gateway nodes, and the non-gateway nodes are controlled to collect data to be transmitted to the off-network gateway nodes through local multi-hop links. The invention realizes the cooperative optimization of the gateway in the off-network communication scene, reduces the energy consumption and improves the backhaul reliability.

Inventors

• LU SHI
• HE YINGJIU

Assignees

• 北京牧家科技有限公司

Dates

Publication Date

20260508

Application Date

20260130

Claims (6)

1. 1. The off-network gateway dynamic election method based on the edge AI and the satellite ephemeris is characterized by comprising the following steps of, Forming a candidate gateway node set in an off-network ad hoc network, collecting multi-source running state information aiming at each candidate gateway node in the candidate gateway node set, and constructing node state feature vectors according to a field arrangement sequence; The method comprises the steps of acquiring multisource operation state information, wherein the multisource operation state information comprises satellite communication observation information, off-network self-organizing network state information, energy state information and operation load information, the satellite communication observation information comprises satellite signal-to-noise ratio and visible satellite quantity, the off-network self-organizing network state information comprises the hierarchical position of candidate gateway nodes in an off-network self-organizing network and channel occupation condition, the energy state information comprises residual electric quantity and energy acquisition rate, and the operation load information comprises the length of a data queue to be transmitted; Based on the node state feature vector, intelligent reasoning is performed on the edge side to generate gateway evaluation indexes, gateway availability constraint is applied to the gateway evaluation indexes by combining satellite ephemeris information, and constrained gateway evaluation indexes are formed; The step of performing intelligent reasoning on the edge side to generate gateway evaluation indexes comprises the steps of inputting node state feature vectors corresponding to each candidate gateway node in the candidate gateway node set into a lightweight intelligent model deployed on the node local in an election period, and performing forward reasoning operation; The lightweight intelligent model is constructed by adopting a lightweight fully-connected neural network structure, the fully-connected neural network consists of an input layer, a plurality of hidden layers and an output layer, the dimension of the input layer is consistent with the dimension of the node state feature vector, and the dimension of the output layer is consistent with the number of operation modes in a discrete action space; in the forward reasoning operation process, a state space is constructed based on node state feature vectors, value evaluation is respectively carried out on each operation mode in the discrete operation space by combining with a reward function, so as to obtain an action value estimated value of each operation mode, and the action value estimated value of an active gateway mode is extracted from the action value estimated value of each operation mode and is used as a gateway evaluation index of candidate gateway nodes in the current election period; performing dynamic off-network gateway election according to constrained gateway evaluation indexes, determining off-network gateway nodes, broadcasting gateway role notification information by the off-network gateway nodes, and controlling non-gateway nodes to collect data to be transmitted to the off-network gateway nodes through local multi-hop links; the dynamic off-network gateway election is executed according to the constrained gateway evaluation indexes, and the determining of the off-network gateway node comprises the steps of selecting a candidate gateway node with the highest constrained gateway evaluation index value as the off-network gateway node by carrying out numerical comparison on the constrained gateway evaluation indexes in the available state; After the off-network gateway node is selected, calculating a satellite communication time window of the off-network gateway node based on satellite ephemeris information, triggering the off-network gateway node to execute satellite communication link establishment and data reporting in the satellite communication time window, controlling the off-network gateway node to enter dormancy in a non-satellite communication time window and waking up according to preheating time, and finishing timing decision of data reporting and dormancy waking up control.
2. 2. The method for dynamically electing off-network gateways based on edge AI and satellite ephemeris of claim 1 wherein forming the set of candidate gateway nodes comprises performing gateway capability screening based on node communication capabilities, presence and power states in the off-network ad hoc network during operation of the off-network ad hoc network to form the set of candidate gateway nodes.
3. 3. The off-grid gateway dynamic election method based on edge AI and satellite ephemeris of claim 2 wherein the applying gateway availability constraints to gateway evaluation metrics in combination with satellite ephemeris information to form constrained gateway evaluation metrics includes, after generating gateway evaluation metrics, calculating spatial position vectors of satellites relative to candidate gateway nodes in an election period from satellite ephemeris information by combining node local geographic positions with node local real-time clocks and calculating satellite elevation angles of satellites at any time in the election period from satellite spatial position vectors; calculating a remaining visible window time for the satellite to remain in a selected period that satisfies a communication condition based on the satellite elevation angle and a minimum elevation angle requirement for allowing the satellite communication to be performed; When the time of the residual visible window is smaller than the duration time required by completing the processes of data collection, satellite communication link establishment and data reporting, or the residual electric quantity of the latest sampling of the candidate gateway node is lower than the safe lower limit electric quantity required by the operation of the gateway role, marking the gateway evaluation index as an unavailable state; When the time of the remaining visible window is not less than the duration time required by completing the processes of data collection, satellite communication link establishment and data reporting, and the remaining electric quantity of the candidate gateway node sampled at the latest time is not less than the safety lower limit electric quantity required by the gateway role operation, the original gateway evaluation index value is reserved, and an available state mark is added to the gateway evaluation index to obtain a constrained gateway evaluation index.
4. 4. The method for dynamically electing off-network gateways based on edge AI and satellite ephemeris according to claim 3 wherein the controlling the non-gateway node to aggregate the data to be sent to the off-network gateway node via the local multi-hop link comprises updating the node role state of the non-gateway node in the off-network ad hoc network after receiving the gateway role notification information, not performing the operations of establishing a satellite communication link and reporting the data in the current election period, and forwarding the data to be sent generated or cached by the node to the off-network gateway node via the local multi-hop communication link in the off-network ad hoc network; The off-network gateway node receives the data to be transmitted forwarded by the non-gateway node, and performs aggregation processing on the received data to be transmitted to form data to be reported.
5. 5. The method for dynamically electing an off-network gateway based on an edge AI and a satellite ephemeris of claim 4 wherein calculating a satellite communication time window for the off-network gateway node based on the satellite ephemeris information includes determining a satellite communication time window from a time start time and a time end time of the satellite communication time window, the time when an elevation angle of the satellite meets a minimum elevation angle requirement for allowing the satellite communication to be performed for the first time, the time start time and the time end time when the elevation angle of the satellite does not meet the minimum elevation angle requirement for allowing the satellite communication to be performed for the satellite communication time window.
6. 6. An off-network gateway dynamic election system based on an edge AI and a satellite ephemeris is characterized by comprising the following steps of, The candidate screening module is used for screening nodes with gateway capability in the off-network ad hoc network to form a candidate gateway node set and constructing node state feature vectors of the candidate gateway nodes; The intelligent evaluation module is used for executing light intelligent reasoning on the edge side based on the node state feature vector, generating gateway evaluation indexes and applying gateway availability constraint; the gateway election module is used for determining off-network gateway nodes according to constrained gateway evaluation indexes and controlling non-gateway nodes to collect data to be sent to the off-network gateway nodes; and the time sequence control module is used for calculating a satellite communication time window based on the satellite ephemeris information and executing data reporting time sequence decision and dormancy wakeup control.

Description

Off-network gateway dynamic election method and system based on edge AI and satellite ephemeris Technical Field The invention relates to the technical field of off-network communication networks, in particular to an off-network gateway dynamic election method and system based on edge AI and satellite ephemeris. Background In application scenarios such as forestry, agriculture, ocean monitoring, mineral exploration, wild animal protection, disaster emergency response, and the like, a large number of intelligent terminals are deployed in an off-network environment lacking cellular communication infrastructure coverage for a long time. The terminal generally builds an ad hoc network by means of a low-power wireless communication technology, and realizes data acquisition and local cooperative transmission among nodes. In order to complete trans-regional data return, the existing system generally introduces a node with satellite communication capability to take on a gateway role, and transmits collected data to a remote server through a satellite communication link, so that an end-cloud communication system is formed to support information acquisition and remote management requirements in an off-network scene. However, the existing method still has two limitations that firstly, the gateway node is selected by depending on static rules or finite state indexes, the change characteristics of node communication conditions, energy states and load conditions under dynamic environments are difficult to comprehensively describe, secondly, satellite communication has obvious timeliness and intermittence, the traditional method generally responds to the link availability change passively after the gateway is selected, the foresight utilization of satellite orbits and communicable time windows is lacking, and the degree of cooperation between gateway decisions and actual satellite communication conditions is limited. Disclosure of Invention The present invention has been made in view of the above-described problems occurring in the prior art. Therefore, the invention provides an off-network gateway dynamic election method based on the edge AI and the satellite ephemeris, which solves the problem of insufficient cooperation between the gateway election decision and the node running state and the satellite communication available condition in the off-network environment. In order to solve the technical problems, the invention provides the following technical scheme: In a first aspect, the invention provides an off-network gateway dynamic election method based on edge AI and satellite ephemeris, which comprises the steps of forming a candidate gateway node set in an off-network ad hoc network, collecting multi-source running state information for each candidate gateway node in the candidate gateway node set, and constructing node state feature vectors according to a field arrangement sequence; Based on the node state feature vector, intelligent reasoning is performed on the edge side to generate gateway evaluation indexes, gateway availability constraint is applied to the gateway evaluation indexes by combining satellite ephemeris information, and constrained gateway evaluation indexes are formed; performing dynamic off-network gateway election according to constrained gateway evaluation indexes, determining off-network gateway nodes, broadcasting gateway role notification information by the off-network gateway nodes, and controlling non-gateway nodes to collect data to be transmitted to the off-network gateway nodes through local multi-hop links; After the off-network gateway node is selected, calculating a satellite communication time window of the off-network gateway node based on satellite ephemeris information, triggering the off-network gateway node to execute satellite communication link establishment and data reporting in the satellite communication time window, controlling the off-network gateway node to enter dormancy in a non-satellite communication time window and waking up according to preheating time, and finishing timing decision of data reporting and dormancy waking up control. The method for dynamically selecting the off-network gateway based on the edge AI and the satellite ephemeris comprises the steps of forming a candidate gateway node set, wherein in the running process of the off-network ad hoc network, gateway capability screening is performed based on node communication capability, online state and electric quantity state in the off-network ad hoc network, and the candidate gateway node set is formed. The off-network gateway dynamic election method based on the edge AI and the satellite ephemeris comprises the steps of collecting multi-source operation state information including satellite communication observation information, off-network self-organizing network state information, energy state information and operation load information, wherein the satellite communication observation information incl