Not logged in.
Quick Search - Contribution
Contribution Details
| Type | Bachelor's Thesis |
| Scope | Discipline-based scholarship |
| Title | Intelligent Framework to Detect Ransomware Affecting Linux-based and Resource-constrained Devices |
| Organization Unit | |
| Authors |
|
| Supervisors |
|
| Institution | University of Zurich |
| Faculty | Faculty of Business, Economics and Informatics |
| Date | 2022 |
| Abstract Text | The Internet of Things (IoT), a network of interconnected devices, has been growing and gaining traction in various industries. This technology can impact our lives while also providing significant economic benefits. For example, crowdsensing platforms such as ElectroSense that use sensor-equipped IoT devices to collect and share spectrum monitoring data have proven efficient, cost-effective, and scalable. However, although these resource-constrained IoT devices provide numerous benefits, they are also vulnerable to cyberattacks. As a result, ransomware could severely threaten the IoT ecosystem. ElectroSense, which employs IoT device sensors, may fall victim to such an attack, resulting in operational problems and sensor data unavailability. Machine and deep learning algorithms using behavioral data have been identified as promising ransomware detection and classification techniques. However, most detection frameworks that utilize these technologies have been developed for Windows-based systems, which generally have more resources than IoT devices. As a result, these solutions may not be well-suited for crowdsensing platforms which utilize resource-constrained components. In addition, while ransomware policies are effective and resource efficient in detecting and classifying ransomware, they do have some limitations. This thesis, therefore, proposes to develop and test a machine and deep learning-based framework that utilizes three different behavioral sources to detect and classify ransomware impacting resource-constrained ElectoSense sensors. This framework will employ an efficient, scalable, and data-protective approach to identify zero-day ransomware attacks and classify various ransomware strains. In addition, real-world ransomware attack scenarios are utilized to test the platform's effectiveness. |
| PDF File |
Download
|
| Export |
BibTeX
|