Edge-Cloud Collaborative IoT System Architecture and its Reliable Transmission Mechanism for Geological Disaster Monitoring

Abstract

Aiming at the critical problems of high data transmission latency, highly vulnerable communication links, and overwhelming processing loads on central servers inherent in traditional geological disaster monitoring systems deployed in complex mountainous environments, an advanced edge-cloud collaborative Internet of Things system architecture alongside a highly reliable heterogeneous transmission mechanism is proposed. First, a comprehensive four-layer Internet of Things architecture comprising the perception, network, platform, and application layers is meticulously designed to establish clear functional boundaries and operational continuity. An innovative edge-cloud collaborative processing mechanism is introduced, wherein distributed edge computing nodes, utilizing ultra-lowpower microcontrollers and capable single-board computers, perform the immediate initial filtering of multi-source sensor data, rigorous outlier removal, and localized threshold-based early warning generation. This decentralization dramatically reduces upstream transmission bandwidth requirements and localized response latency. Concurrently, the centralized cloud platform leverages sophisticated multi-sensor data fusion algorithms and complex physical models to conduct in-depth historical data mining and comprehensive global risk assessments. Second, targeting the fundamental pain point of extreme network instability in rugged topographies, a reliable transmission network based primarily on heterogeneous Low-Power Wide-Area Networks, incorporating both Long Range and Narrowband Internet of Things technologies, is engineered with multi-communication backup mechanisms and dynamic failover logic. Additionally, the lightweight Message Queuing Telemetry Transport protocol, combined with localized payload encryption and adaptive data compression algorithms, is deployed at the transmission layer. This strategic protocol optimization ensures secure, high-throughput, and energy-efficient data transmission in the harshest of environmental conditions. Experimental results demonstrate that the proposed adaptive hybrid transmission strategy achieves packet delivery ratios of 99.5%-99.9% over transmission distances of 2-10 km. Meanwhile, the edge-cloud collaborative architecture reduces the end-to-end response time to $0.97-3.65 ~\mathrm{s}$ across the evaluated event types. These results verify that the proposed architecture can effectively improve transmission reliability and real-time warning performance in complex mountainous environments.