Mastergradsoppgave   2017

Gebrie, Mattias Tsegaye

Publikasjonsdetaljer

Veiledet av:

Abie, Habtamu; Ernesto, Prinetto Paolo; Farulla, Giuseppe Airò

Utgiver:

Politecnico di Torino

Antall sider: 77

Health care is one of the primary beneficiaries of the technological revolution created by Internet of Things (IoT). In the implementation of health care with IoT, wireless body area network (WBAN) is a suitable communication tool. That being the case security has been one of the major concerns to efficiently utilize the services of WBAN. The diverse nature of the technologies involved in WBAN, the broadcast nature of wireless networks, and the existence of resource constrained devices are the main challenges to implement heavy security protocols for WBAN. In this paper we develop a risk-based adaptive authentication mechanism which continuously monitors the channel characteristics variation, analyzes a potential risk using naive Bayes machine learning algorithm and performs adaptation of the authentication solution. Our solution validates both the authenticity of the user and the device. In addition we evaluate the resource need of the selected authentication solution and provide an offloading functionality in case of scarce resource to perform the selected protocol. The approach is novel because it defines the whole adaptation process and methods required in each phase of the adaptation. The paper also briefly describes the evaluation case study - Smart Home eHealth.