ConQeng: A Middleware for Quality of Context Aware Selection, Measurement and Validation

Abstract

A set of quality metrics (e.g., timeliness, completeness) together represent the Quality of Context (QoC); their values determine the usability of context to context consumers (IoT applications). Therefore, obtaining adequate ‘QoC from the context providers (context sources) represents a significant research challenge. This paper presents a framework called conQeng that addresses such a challenge through novel approaches in QoC-aware selection, QoC measurement and validation. ConQeng selects the potential context providers that deliver an adequate QoC during runtime, assesses their performance - for further selection, and transfers QoC-assured context to the context management platforms (CMPs). We have implemented conQeng in a simulated scenario involving autonomous cars, marketing service agencies as context consumers, and thermal and video cameras as context providers. The results demonstrate that it outperforms three heuristic approaches in reducing context acquisition cost and improving effectiveness and performance efficiency while obtaining adequate QoC.