High-performance scheduling model for multisensor gateway of cloud sensor system-based smart-living

In cloud sensor system-based smart-living applications, large-scale distributed sensors may be deployed to collect information and report to the manipulator and the cloud. A gateway is often employed as infrastructure in this scenario, acting as the data collector, the relay and the agency of the initial multisensor fusion, and thus must be able to handle as many concurrent requests as possible from diverse sensors of different vendors. This study proposes a high-performance scheduling model with a cloud-supported caching mechanism for the gateway of the cloud sensor system-based smart-living. Scheduling and caching optimization are performed by 0–1 programming combined with the periodic task models. Correlation analyses of the simulated results determine the most effective factors to the performance, and the performance tests with the selected factors show that a gateway with 2.4 G-uniCPU/4 G-memory/300 G-harddisk can support the system with one million sensors registered in the cloud and 5000 concurrent live sensors through it, with a 25× gain throughput improvement compared to the traditional application-type scheduling.