Abstract
The advent of pervasive ubiquitous computing and advancement of wireless
communication technologies has resulted in the proliferation of innovative
mobile computing devices like tablets and smartphones. In
consumer market and business community worldwide, smartphones have
become the most reliable portable devices for Internet connectivity and
sensitive data storage. As smartphones are becoming the core delivery
platform for ubiquitous “connected customer services” paradigm;
security threats and concomitant risks are also growing proportionally.
Recent reports from security vendors highlight this emerging challenge.
Additionally, smartphones pose limitations for security solution architects,
such as limited computing power, memory, battery and peripherals
etc. This makes the desktop security countermeasures infeasible for
smartphone devices. Some known anti-malware commercial products
for smartphones – by top ranked security vendors – are signature based
and require continuous updating for latest malware detection. Moreover,
these products are unable to detect the zero-day and polymorphic
malwares for smartphones. Therefore, we argue that the domain of
non-signature based anti-malware solutions for smartphones is open for
research.
In this dissertation, a novel security framework is proposed for malware
detection on Linux based computers and smartphones using different
data mining approaches.
more here.......http://nexginrc.org/Publications/pub_files/Farrukh_thesis_v2.pdf
The advent of pervasive ubiquitous computing and advancement of wireless
communication technologies has resulted in the proliferation of innovative
mobile computing devices like tablets and smartphones. In
consumer market and business community worldwide, smartphones have
become the most reliable portable devices for Internet connectivity and
sensitive data storage. As smartphones are becoming the core delivery
platform for ubiquitous “connected customer services” paradigm;
security threats and concomitant risks are also growing proportionally.
Recent reports from security vendors highlight this emerging challenge.
Additionally, smartphones pose limitations for security solution architects,
such as limited computing power, memory, battery and peripherals
etc. This makes the desktop security countermeasures infeasible for
smartphone devices. Some known anti-malware commercial products
for smartphones – by top ranked security vendors – are signature based
and require continuous updating for latest malware detection. Moreover,
these products are unable to detect the zero-day and polymorphic
malwares for smartphones. Therefore, we argue that the domain of
non-signature based anti-malware solutions for smartphones is open for
research.
In this dissertation, a novel security framework is proposed for malware
detection on Linux based computers and smartphones using different
data mining approaches.
more here.......http://nexginrc.org/Publications/pub_files/Farrukh_thesis_v2.pdf