Data centers underpin modern digital infrastructure but face key challenges in networking, particularly in programmability, performance, and security. Technologies like OpenvSwitch, P4 and other Software-Defined Networking (SDN) paradigms increase the flexibility of packet processing, yet P4 performance on network adapters is not always deterministic, due to network pipeline design, Match-Action Tables (MATs) and their entries, as well as traffic during runtime. Achieving consistent line-rate performance requires both efficient pipeline design and runtime optimization. At the same time, securing data across network slices and domains, especially in emerging 6G and multi-tenant environments, demands quantum-resilient cryptographic algorithms. This PhD thesis addresses these issues by proposing P4 runtime optimization techniques, analyzing NVIDIA’s DOCA Pipeline Language (DPL), and introducing SNSVS and QRoNS, which are frameworks enable end-to-end quantum-secure communication using Post-Quantum Cryptography through IPsec tunnels.

After the defence there will be a small reception at Frederikskaj 12, 3rd floor, room 3.49a