Adaptive data caching for provisioning sensors-as-a-service

This work focuses on designing an optimal, and adaptive data caching mechanism to be implemented within sensor-cloud environment. Data caching within sensor-cloud is already explored in a previous work. However, the work lacked timeliness of data delivery, i.e., the response time of a query was large due to repeated caching. Further, the design of the work did not consider incoming sensor data within an unstable environment. To address these difficulties, we propose a revised scheme that modifies the existing caching architecture. This work models the data caching to be followed with the External Cache (EC) and the Internal Cache (IC). Results show that the new model considers both stable and unstable environment and accurately estimates the change of environment over time. Further, the experimental results depict that the timeliness of data is improved to 93.4% and the queries undergo comparatively lower response time thereby, justifying the real-time applicability of the proposed scheme.