Search for: All records

Significant growth of internet applications in recent years has raised a lot of challenges to networks. One of the important applications is smart and connected health (SCH), which utilizes sensing, communication networks and artificial intelligent (AI) techniques to offer healthcare services to the users. In SCH applications, Wireless Body Area Networks (WBANs) consisting of a group of Lightweight and wearable devices designed for use within the proximity of the human body, is a key infrastructure. In this short paper, we discussed the possibility of exploring AI techniques for WBANs to improve network performance and enhance health services. In addition, we present the literature review of AI driven networks for SCH, its related challenges and future directions. 

Securely growing or de-growing nodes is a mandatory requirement to manage Wireless Body Area Networks (WBANs). This requirement raises significant challenges in node authentication, backward node authentication, initial node configuration, and node de-growth. Unlike the traditional approaches using pre-stored secrets or relying on special authentication hardware, we explore the characteristics of WBAN and wireless signal to develop an efficient scheme for adding/removing WBAN node securely and effectively. The major idea of the proposed scheme is to construct a 'virtual' dual-antennae proximity detection system by fully utilizing the existing legitimate nodes and the behavior of human body. We built a system prototype on wireless devices and verified our scheme through experiments. In addition, a data mining (clustering) algorithm is also applied to successfully detect newly joined legitimate node and identify potential attackers.