Explicit context matching in content-based publish/subscribe systems.

Area of Science:

• Computer Science
• Distributed Systems
• Networking

Background:

• Most publish/subscribe (P/S) systems are context-agnostic, failing to leverage dynamic context for improved performance in areas like wireless sensor networks (WSNs) and IoT.
• Existing P/S models often treat context as regular application data, missing opportunities for optimization.

Purpose of the Study:

• To introduce SilboPS, a novel context-aware P/S model designed for efficient context management.
• To minimize network overhead in distributed systems with frequently changing contexts, such as mobile ad hoc networks (MANETs).
• To enhance the sharing and utilization of sensor data from WSNs for context-aware applications.

Main Methods:

• Developed a novel P/S model (SilboPS) that explicitly manages context.
• Focused on minimizing network overhead in dynamic environments with recurrent context changes.
• Implemented contextual scoping within the filtering process to decouple subscriptions from their production context.

Main Results:

• Empirically demonstrated significant reduction in (un)subscription cost per node.
• Showcased improvement in the global performance and throughput of the broker network.
• Maintained the cost of topology changes comparable to existing SIENA-like systems.

Conclusions:

• SilboPS offers an efficient solution for context-aware data sharing in WSNs and IoT applications.
• Decoupling subscriptions from context and using contextual scoping optimizes network resource utilization.
• The model effectively enhances context awareness and system performance in dynamic distributed environments.