Enabling Large-Scale IoT-Based Services through Elastic Publish/Subscribe.

Area of Science:

• Cloud Computing
• Internet of Things (IoT)
• Distributed Systems

Background:

• Internet of Things (IoT) services require scalable cloud infrastructure to handle continuous data streams from sensors.
• Elastic scaling in cloud environments can disrupt data flow, impacting the reliability of IoT services.
• Existing publish/subscribe systems may not adequately support the dynamic and context-aware nature of IoT data.

Purpose of the Study:

• To present an algorithm for elastic cloud scaling in IoT infrastructure.
• To ensure uninterrupted data reception for IoT service users during scaling operations.
• To describe the implementation and evaluation of this algorithm within the E-SilboPS system.

Main Methods:

• Developed an algorithm for elastic cloud infrastructure to support IoT.
• Implemented the algorithm in E-SilboPS, an elastic content-based publish/subscribe (CBPS) system.
• Evaluated the scale-out process using quantitative measurements.

Main Results:

• The algorithm provides uninterrupted service during scaling operations through transparent state repartitioning.
• Scaling operations are time-bounded and dependent on state partition dimensions.
• E-SilboPS successfully supports context-aware sensing and communication in IoT services.

Conclusions:

• The developed algorithm and E-SilboPS system effectively address the challenge of maintaining continuous data flow in elastic IoT cloud infrastructures.
• The system offers middleware for dynamic dashboards and monitoring panels in IoT services.
• This advancement enhances the context-aware content-based features of E-SilboPS for IoT applications.