An Adaptive Resource Allocation Approach Based on User Demand Forecasting for E-Healthcare Systems

The Internet of Things (IoT) and Mobile Edge Computing (MEC) hold great potential in e-Health. Wireless Body Area Network (WBAN) as one of the primary Internet of Things (IoT) provides real-time and continuous healthcare monitoring and has been widely deployed to improve people's quality of life. MEC can support WBAN better with reliable connection reliability and high computing performance. However, with the rapid development and increasing number of WBAN, user mobility and uncertainty make it greatly challenging to allocate resources in MEC for achieving low latency and high resource utilization. Existing bottlenecks of allocation strategy include high computational complexity, weak robustness, and low quality of service (QoS). It is vital to develop a strategy that can adaptively adjust resource allocation, detect and respond to load changing timely. To this end, we propose an adaptive resource allocation scheme based on demand forecasting. Specifically, the proposed algorithm employs forecasting model to extract user behavior characteristics, quantifies the forecast results reasonably by introducing queuing theory, which provides a basis for the matching of resources and users. Moreover, we propose the dynamic communication and computing resource allocation scheme to pre-allocate resources and release unnecessary resources. The simulation evaluations show that the proposed model achieves better resource utilization (at least 25% improvement) than existing methods, and is robust in guaranteeing user service quality.