Abstract:
As the usage of smart-devices is increasing, security threats affecting the confidentiality,
integrity, and privacy of such devices are briskly emerging due to the rapid growth
of malware. Mobile security suites exist to defend devices against malware and other
intrusions. However, they require extensive resources which is a constraint of the device
itself. In this thesis, we address the problem of intrusion detection for both smartdevices
and vehicles while taking into account the devices’ limited resources such as
energy, CPU usage, and Internet connectivity via intelligent offloading. We provide
a mobile and vehicular ad-hoc cloud based intrusion detection framework that takes
advantage of Wi-Fi Direct to allow connectivity, sharing resources, and integrating security
as a service with or without the availability of an Internet connection. The frame-framework
leverages intelligent offloading via an Intelligent Offloading Distributor module
that outputs the optimal offloading decision and distribution. The experiments demonstrate
various improvements achieved by our framework. Based on our IOD module,
our approach is capable of significantly reducing energy consumption, execution time,
and number of selected computational nodes.
