GAP4NSH: a genetic service function chaining with network service header for P4-based software-defined networks

This study proposes a service function chaining (SFC) management system architecture that is based on a genetic algorithm and network service header encapsulation that is called GAP4NSH. GAP4NSH involves proof-of-concept implementation and optimization of SFC deployment management with the network service header (NSH) and SFC-Proxy in a P4-based software-defined network. GAP4NSH consists of a genetic algorithm-based SFC (GASFC) algorithm, a P4NSH SFC deployment mechanism, and an optimized NSH routing strategy. GASFC can efficiently allocate network services to various SFCs in complex network topologies and dynamically calculate suitable resources on the basis of various SFC paths. P4NSH contains an optimized NSH rule strategy that can operate in two modes, and it can improve the memory usage efficiency of P4 switches. Our proof-of-concept implementation is validated by seven-scenario experiments. In comparison with alternative advanced service functions deployment methods, the proposed GAP4NSH scheme demonstrates a substantial reduction in the time for service functions deployment of the service function chain by 83.67% and also significantly reduces the network management traffic within the SDN P4 controller by 94.23%. These improvements contribute to a significant enhancement in the overall effectiveness of SDN network management. The GAP4NSH is compatible with all traditional network services, so it can ease technical obstacles to developing SFC in software-defined networks.