Abstract

Flex Net Sim is a robust, high-performance, open-source C++ library designed to simulate resource allocation heuristics in elastic optical networks. Based on an event-driven simulation paradigm, the library has been purpose-built to address routing, modulation format, and spectrum allocation problems, as well as other related optimization challenges. With Flex Net Sim, researchers can focus solely on heuristic code development without requiring extensive programming expertise or complex configurations. Additionally, Flex Net Sim includes native tools for statistical analysis, including confidence interval estimation. To demonstrate its capabilities, a comprehensive benchmarking study was conducted using six allocation algorithms: first fit (FF); exact fit (EF); random fit (RF); best fit (BF); least loaded (LL); and proactive, fragmentation-aware spectrum management (PFA), across 105 real network topologies. The results indicate that LL consistently achieves the lowest blocking probability, while RF performs the worst in most scenarios. The library is fully extensible, actively maintained, and released under the MIT license, with source code available in a public repository.