Mobile App Profiling for Malicious Activity Detection

  • Mtetwa, Nhamoinesu (PI)

    Project Details

    Description

    The aim of this proposal is to productise technology aimed at fundamentally securing mobile platforms against potential threats from malicious software applications. Support is sought to integrate research and expertise at Glasgow Caledonian University from the areas of malware analysis, mobile forensic analysis and machine learning to enable the development of a novel profiling system for the dynamic detection of malicious activity on mobile-platform applications.

    If successful, the software solution will lead to the formation of a Scottish high growth company specifically to address the needs of organisations with rigorously-enforced security audit and compliance regimes such as the financial sector, security services, direct marketing industries, health care, and Government interfaces with the private sector. In addition, a security-as-a-service business model is envisioned to target individual security-conscious users.

    Distinguishing between benign and malicious activity on mobile platforms is a difficult to solve problem due to low-power availability and limited resources. The proposal aims to integrate capture of low-level activity on the device, including permissions, systems call sequences, API call sequences, memory access, inbound and outbound network activity, with novel machine-learning-based analysis in order to verify benign device performance.

    Our solution is innovative in that it will combine dynamic mobile App profiling on the device and also further analysis in the cloud. This is a smart combination of edge computing and cloud computing. This solution benefits from our research in mobile forensics, malware analysis, networking systems, machine learning, cloud computing and embedded systems.
    StatusFinished
    Effective start/end date12/02/1812/04/18

    Funding

    • Scottish Government: £11,112.00

    Fingerprint

    Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.