Traditional data center networks rely on fixed topologies and manual configurations, making it challenging to respond promptly to fluctuating workloads and traffic patterns. In contrast, reconfigurable Data Center Networks (RDCNs) represent a significant departure from traditional data center networks in terms of flexibility, scalability, and adaptability . Unlike conventional static networks, RDCNs embrace dynamic and programmable architectures, enabling rapid adjustments to network configurations and resource allocations based on real-time traffic demands. RDCNs can dynamically reconfigure their routing paths, apply load balancing techniques, and reroute traffic in the event of link failures, resulting in enhanced resilience and reduced downtime. This level of adaptability also enables RDCNs to scale more efficiently, accommodating the exponential growth of data and the increasing demands of modern applications. Ultimately, reconfigurable data center networks offer the agility and efficiency needed to meet the ever-evolving requirements of contemporary data centers, providing improved performance and cost-effectiveness compared to their static counterparts.
In this thesis, the student will analyze existing problems in RDCNs and design solutions (and/or evaluate existing ones) to some of those problems. The focus can be on the topology, routing, or network protocols, depending on student’s interests and expertise. Implementation and evaluation of those solutions will be carried on through simulations.
Approximate composition: 25% State of the art analysis, 35% Theory/Design, 40% Implementation/Experiments