Edge iot asynchronous agent orchestration method and system
By extending the address-event representation frame sequence and asynchronous handshake protocol, the task graph is dynamically segmented into subgraphs and the target intelligent agent is selected, which solves the problem of task-level dynamic routing across heterogeneous edge nodes, realizes low-latency and low-power task orchestration and collaborative execution, and meets the real-time and energy efficiency requirements of edge IoT.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SCHOOL OF SOFTWARE ZHEJIANG UNIV (NINGBO) MANAGEMENT CENT (NINGBO SOFTWARE EDUCATION CENT)
- Filing Date
- 2026-06-22
- Publication Date
- 2026-07-21
AI Technical Summary
Existing technologies cannot effectively solve the problem of task-level dynamic routing and runtime rerouting across heterogeneous edge nodes, resulting in excessive communication latency and power consumption, which cannot meet the millisecond-level real-time requirements and milliwatt-level power consumption budget of edge IoT.
Extended address-event representation is used to serialize task pulse events in frames. Combined with a global asynchronous local synchronization strategy and a four-phase asynchronous handshake protocol, the task directed acyclic graph is dynamically divided into subgraphs. The target asynchronous agent is selected by the principle of weighted minimum cost, realizing task-level dynamic routing and rerouting across heterogeneous edge nodes. Vector clocks are used to maintain causal order relationships for result aggregation.
It significantly reduces communication latency and power consumption, enables dynamic rerouting and loosely coupled collaboration across heterogeneous edge nodes, meets the real-time and energy efficiency requirements of edge IoT, and demonstrates excellent task orchestration and collaborative execution capabilities in drone swarms, medical sensors, and industrial IoT scenarios.
Smart Images

Figure CN122437807A_ABST