Big Data: Infrastructure and Analytics (BDIA)

Research Problem and Purpose

The data revolution introduced by the Internet companies from the early 2000s has now been amplified by the wide adoption of mobile devices. In almost every domain data is available in a high volume and a high variety, and often accumulated at a high velocity. This phenomenon is now acknowledged as Big Data. The rise of big data introduces new challenges in data manipulation and access.

The aim of this research focus area (RFA) is to investigate and devise new approaches for storing, exploring and accessing data faster and in an efficient manner. In short, our contributions include data infrastructure (both the tools and the network), advanced fault-tolerance mechanisms to guarantee robustness in such systems and finally more analytics to further understand the data at hand.

This RFA includes the Dependable Advanced Network Systems cluster. Our endeavour is organised around three essential themes. These are: (i) Smart Infrastructures; (ii)Dependable Systems; (iii) Advanced Analytics.

  • Smart Infrastructures
    One of the core challenges introduced by the Big Data revolution is to provide storage techniques to cope with the high volume, variety and velocity of data. In recent years we have seen a new wave in data management systems with the introduction of NoSQL1 data stores to complement the traditional relational data management systems. However, the challenge being faced goes far beyond data management systems; the data is usually not only stored in its input format but requires additional internal transformations to support further exploration. These transformations can occur in the form of a batch, real-time or stream processing.
    Thus, we wish to borrow ideas from system virtualization and service orientation to provide dynamic infrastructures that can address the need for heterogeneity both for the data and the set of operations applied to the data. Furthermore, in order to reduce the latency in the complex infrastructures we seek to provide, there is a need for a suitable network infrastructure be achieved through a proper management of bandwidth and hence efficient schemes of quality of service (QoS) provisioning in the system. Despite the multitude of frameworks for QoS provisioning at various layers of the network protocol stack there still exist many challenges to be tackled both for wired and wireless networks. Our work in infrastructure seeks to advanced research contributions in this area as well. Knowledge Areas: Distributed Systems, Databases, Computer Networks, Virtualization.
  • Dependable Systems
    Traditionally, the key properties of systems built for Big Data include fault-tolerance and resilience. These properties require additional architectural mechanisms beyond replication and process coordination. Here we wish to introduce the core principles of Dependable Systems. Generally, such systems are aimed at improving the livelihood of society by creating systems that are safe, reliable, available and maintainable. By introducing dependability analysis we wish to not only make our systems robust but also to minimise their cost (e.g., cost spent on system design and or maintenance). The research under this theme will develop sophisticated techniques to advance the robustness of our systems. Knowledge Areas: Dependability Analysis, Fault-Tolerance, Modelling, Optimization.
  • Advanced Analytics
    The deluge of data accumulated with Big Data proves useful when it can help answer not only the questions of the moment but also predict the future. To achieve the latter, advanced analytics techniques are needed. These are computationally intelligent methods and techniques to further understand data and discover new behavioural patterns. Further understanding data is paramount as it plays a key role in decision making (policies, system design, etc.) in governments and industries. Under this theme we will explore various techniques in Machine Learning as well as advanced search techniques using genetic algorithms, simulated annealing, hill climbing, Tabu search, ant colony, etc. Knowledge Areas: Machine Learning, Search Algorithms, Optimization, Data Mining.

Socio Economic and Academic Rationale
Recently the Government of Namibia has pledged to put forth an action plan towards the development of a knowledge economy. A knowledge economy implies that key decisions are made based on data. As such the Government of Namibia needs to equip itself with cost-effective mechanisms to host large volumes of data but also advanced techniques to understand the data and mine fresh information from it. Our efforts in this RFA will pave the way towards this goal. From an academic point of view the topics we touch on will help sharpen the skills of our students and build capacity among faculties. Furthermore, this will eventually lead to the development of new curricula suitable for the 21st century economy.

Prof Dharm Singh Jat (
Prof Josse Quenum (


Dependable and Secure  Advanced Network  Systems - Prof Dharm Singh Jat

Faculty Members:

  • Prof Dharm Singh Jat
  • Prof Jose Quenum


  • Ms Valerie Garises
  • Mr Alexander B Shipena
Back to Top