I currently supervise the following postgraduate students:
Enock Mbewe (PhD Candidate) : Cost-aware Security Decision Model
This project investigates and implements mechanisms for enhancing security of community network infrastructure with configurable secure QoS. In particular, this study will investigate the use of a novel, cost-aware Internet security decision model to allow users, especially those with limited computing skills, to easily configure security options that can map to complex Internet security mechanisms in order to achieve Confidentiality, Integrity, Authentication and Privacy.
Luqmaan Salie (MSc student) : SDN Traffic Engineering using Segment Routing and DNS
This project explores a Software Defined Networking (SDN) and Segment Routing (SR) solution to improve performance and QoE in national research and education networks (NRENs), focusing on compartmentalizing routing and flows based on traffic types. Solutions will be tested in a virtual environment based on a scaled-down emulation of SANReN, the nation-wide NREN currently in place in South Africa, connecting universities, science institutions, and overseeing science projects such as the Square Kilometre Array. The SDN controller will be incorporated with a DNS resolver to improve optimize network path creation.
Chikomborero Mwenje (MSc student): Content caching in Community Networks
The project explores cache placement strategies in community networks, and in particular, to investigate vital network metrics to consider when deciding on cache placement in small cell base station architecture called cloudlets. This work is motivated by the understanding that an increasing percentage of user devices run content-intensive applications, such as social media applications, resulting in an increased demand for content and increased delays in content delivery. Caching content in cloudlets aims to reduce traffic and cellular network overload.
Chiratidzo Matowe (MSc student): Using Deep Learning to Classify Network Traffic in a Community Network
This project implements a Deep Learning (DL) classifier for purposes of quality of service and traffic engineering in community networks. This research explores and builds a DL model for real-time fine-grained traffic classification using minimal computational resources. In order to respond to the network resource and security challenges, this research will investigate how to select relevant detection features in a federated network measurements environment.