Parallel Processing of Sensor Data in a Distributed Rules Engine Environment through Clustering and Data Flow...

Area of Science:

• Computer Science
• Internet of Things
• Smart Cities

Background:

• Smart buildings and cities utilize sensor networks for real-time data acquisition and processing.
• Scalable data handling is crucial for maximizing system throughput in these environments.
• Existing systems face challenges in managing large volumes of sensor data efficiently.

Purpose of the Study:

• To propose a scalable model for handling sensor data in smart environments.
• To develop and evaluate parallel computing methods for efficient data flow management.
• To investigate the use of open-source cloud solutions for system management.

Main Methods:

• A weighted dependency graph model for abstracting information flow.
• A parallel k-means clustering algorithm variation for data flow reconfiguration.
• A custom genetic algorithm for optimizing data flow.
• Simulation-based evaluation of proposed algorithms.

Main Results:

• The proposed algorithms significantly reduce rule processing times.
• The methods provide an efficient solution for increasing system scalability.
• Demonstrated seamless integration of increased sensor numbers with resource optimization.

Conclusions:

• The developed parallel computing algorithms offer an efficient solution for smart environment data management.
• These methods enhance the scalability and performance of sensor networks in smart buildings and cities.
• Open-source cloud solutions combined with these algorithms improve overall system efficiency.