Design for energy-efficient and reliable fog-assisted healthcare IoT systems

Cardiovascular disease and diabetes are two of the most dangerous diseases as they are the leading causes of death in all ages. Unfortunately, they cannot be completely cured with the current knowledge and existing technologies. However, they can be effectively managed by applying methods of continuous health monitoring. Nonetheless, it is difficult to achieve a high quality of healthcare with the current health monitoring systems which often have several limitations such as non-mobility support, energy inefficiency, and an insufficiency of advanced services. Therefore, this thesis presents a Fog computing approach focusing on four main tracks, and proposes it as a solution to the existing limitations. In the first track, the main goal is to introduce Fog computing and Fog services into remote health monitoring systems in order to enhance the quality of healthcare. In the second track, a Fog approach providing mobility support in a real-time health monitoring IoT system is proposed. The handover mechanism run by Fog-assisted smart gateways helps to maintain the connection between sensor nodes and the gateways with a minimized latency. Results show that the handover latency of the proposed Fog approach is 10%-50% less than other state-of-the-art mobility support approaches. In the third track, the designs of four energy-efficient health monitoring IoT systems are discussed and developed. Each energy-efficient system and its sensor nodes are designed to serve a specific purpose such as glucose monitoring, ECG monitoring, or fall detection; with the exception of the fourth system which is an advanced and combined system for simultaneously monitoring many diseases such as diabetes and cardiovascular disease. Results show that these sensor nodes can continuously work, depending on the application, up to 70-155 hours when using a 1000 mAh lithium battery. The fourth track mentioned above, provides a Fog-assisted remote health monitoring IoT system for diabetic patients with cardiovascular disease. Via several proposed algorithms such as QT interval extraction, activity status categorization, and fall detection algorithms, the system can process data and detect abnormalities in real-time. Results show that the proposed system using Fog services is a promising approach for improving the treatment of diabetic patients with cardiovascular disease.