A Service Discovery Solution for Edge Choreography-Based Distributed Embedded Systems

Area of Science:

• Computer Science
• Distributed Systems
• Internet of Things (IoT)

Background:

• Edge computing architectures in IoT systems often face challenges with service discovery and dynamic resource management.
• Choreography-based distributed embedded systems require robust mechanisms for service integration and load balancing.

Purpose of the Study:

• To develop and evaluate a service discovery solution for edge choreography-based distributed embedded systems.
• To enhance IoT middleware compatibility with existing messaging protocols and support dynamic system changes.

Main Methods:

• Implemented a broker service for service discovery supporting regular expressions (Regex) in message scope.
• Integrated three message passing models into the choreography middleware.
• Compared CPU usage and latency between request-response and data-centric configurations, and analyzed Regex impact on system resources.

Main Results:

• The proposed solution facilitates blind hot plugging of new machines and aids in service load balancing.
• Analysis of Regex interpreter latency and its impact on CPU and memory consumption was performed.
• Performance was evaluated in both request-response and data-centric scenarios.

Conclusions:

• The developed service discovery mechanism effectively supports dynamic service management in IoT edge environments.
• The use of Regex in message scope offers a flexible approach for service matching in distributed systems.
• The solution demonstrates potential for improving the efficiency and adaptability of edge choreography-based systems.