FC-CACPHS: fog-cloud assisted context-aware framework for cyber-physical healthcare system

The advancements in cyber-physical systems (CPS) have brought significant changes to the healthcare industry, especially in the exchange of information. Medical CPS integrates smart data collection devices with cyberspace components for data analytics and decision making. However, this integration poses challenges such as event location, computation overhead, and ubiquitous access. To address these challenges, a scalable, context-aware multilayered MCPS framework based on the fog-cloud paradigm is proposed. The proposed naïve Bayes classifier is experimented with in simulated settings. The results of the naïve Bayes classification component are also compared with the results obtained using several state-of-the-art classification algorithms namely artificial neural networks (ANN), decision trees (DT), and k-nearest neighbour (k-NN). The results reveal that the naïve Bayes classifier outperforms other classification algorithms with the resulting accuracy of 96.7% and specificity, sensitivity, and f-measure of 97.5%, 95.6, %, and 92.86% respectively. The results show that it performs better than these algorithms on typical benchmark datasets.