Adaptive Context Monitoring Framework for Enhancing Caching Efficiency in Context Management Platforms

Abstract

As the Internet of Things (IoT) continues to expand, the volume of data generated by IoT devices and the demand for IoT applications is increasing exponentially. These applications critically rely on real-time context reasoned from IoT data for effective decision-making and actuation, thereby making the accessibility of this context crucial. This study introduces a novel adaptive Context Monitoring Framework (CMF) for enhancing context caching efficiency in Context Management Platforms (CMPs) to better support the near real-time needs of IoT applications. Our proposed framework integrates two novel components: the Context Attributes Prioritisation Engine (CAPE), which prioritises and assigns weights to the context, and the Adaptive Context Management Engine (ACME), which dynamically adjusts thresholds for each context based on incoming query volumes and context cache performance. Combined, our hybrid approach ensures timely updates of context within the cache while also serving context in real-time (reducing any query response latency). Our approach is effective for dynamic changes in an IoT environment through the adaptive approach of continuously monitoring and updating the cached context. We implemented the proposed adaptive framework using a CMP namely the context-as-a-service (CoaaS) platform and evaluated it using real-world datasets obtained from a smart city application. A thorough experimental evaluation demonstrated a marked improvement in cache efficiency, achieving a 90% cache hit rate and reducing the cache expiry ratio to 5%.

Keywords

• CAPE
• CoaaS
• context freshness
• context caching
• hybrid approach
• IoT