Multi-objective Optimization Algorithm based on PTN Network Index System

Packet Transport Network (PTN), as an efficient transmission network technology in mobile communications in the era of big data, is used by more and more communication operators. With the rapid increase in the number of users and the continuous decline in revenue per user, the existing PTN network must be optimized in all aspects. In the PTN network, the optimization of one indicator often affects other indicators.This paper proposes a multi-objective optimization method based on linear weighting to optimize the two major indicators in the PTN network. The first indicator is the active and standby routing rate using the Label Switching Path (LSP) technology. The second indicator is the link bandwidth occupancy rate after adding the committed information rate (CIR).The method firstly establishes a multi-objective optimization mathematical model, converts multi-objective optimization into single-objective optimization, and then uses the improved Top K shortest path (KSP) algorithm to optimize LSP active and standby routing services and CIR Services with excessive bandwidth usage are switched to reduce the number of services with high bandwidth usage and the same route. The experimental results show that the multi-objective optimization method has been improved from the initial score of 82.1484 to 98.5346, achieving the overall optimal result. The LSP indicator has been improved from 74 to 96.72, and the CIR has been improved from 88.1 to 99.86, reaching the multi-objective overall best, single objective is relatively optimal. When the existing path resources cannot solve the problem, the solution will be optimized by adding resources, which not only enhances the security of the PTN network but also improves the utilization of network resources.