Physical-aware long reach PON planning

Abstract

To solve the planning in long reach PON, we develop a mathematical model to capture the internal constraints in cascade splitter scenarios. The model does not focus on the maximal transmission distance, but also takes into account other physical constraints (e.g. the power budget) and management requirements. Through analyzing the PON solution structure, a two-stage genetic algorithm with Prüfer sequence is proposed to solve this NP-hard planning problem. Prüfer sequence is used to keep the tree structure of PON during the optimization process. The correctness of the proposed algorithm is checked by the enumeration method. The impacts of ONU distribution and power budget on the cost and graph characteristics are studied as well. The simulations show that the ONUs with boundary distribution has a higher average cost than those with cluster distribution. In addition, as the power budget reduces, the average cost increases and the splitter connections tend to be tree topologies with more branches. B Rentao Gu