Q-SR: An extensible optimization framework for segment routing

Segment routing (SR) combines the advantages of source powered by software-defined networking (SDN) paradigm and hop-by-hop in legacy network infrastructure. The recent applications SR multi-domain service function chaining (SFC) entail efficient use multiple segments. However, because computation inefficiency, it is nearly impossible to accurately evaluate whether what extent various types networks will benefit from with segments using conventional approaches. In this paper, we propose a flexible Q-SR framework as well its formulation order fully explore potential an algorithmic perspective. highly extensible design algorithms that can be adapted topologies traffic matrices. For offline setting, develop polynomial time approximation scheme (FPTAS) which find (1+ω)-approximation solution for any specified ω>0 size. To best our knowledge, proposed FPTAS first algorithm compute arbitrarily accurate solution. online primal–dual proven O(1)-competitive violates link capacities factor O(logn), where n node number network. algorithms, prove theoretical performance bounds conduct extensive simulations on synthetic realistic analyze related parameters validate efficiency both scenarios.