A Multi-Dimensional Parameters Evaluation-Based Traffic Balancing Scheme in Optical Networks

Xiaobo Li,Guoli Feng,Run Ma,Yuhang Liu,Xinnan Ha,Weibin Wei,Ning Wang,Xiaosong Yu,Yongli Zhao
DOI: https://doi.org/10.1117/12.2624709
2022-01-01
Abstract:With the rapid development of optical networks, traditional resource allocation methods face challenges raised by network services with intensive distribution and differentiated sensitivity on time delay. Based on the real-time services distribution in the network, the traffic load balancing method dynamically schedules arriving requests at the control layer and evenly distributes it to multiple servers or network links to avoid bottleneck links and achieve balanced network resource utilization and services distribution. This paper proposes a traffic load balancing method based on time/wavelength/space evaluation on the coefficient of variation in optical networks, the purpose is to solve the game relationship between time/wavelength/space resources as a multi-dimensional network scenario. It is kind of a dynamic traffic load balancing method, assessing the degree of network equilibrium according to the conditions of network resources distribution. The routing allocation scheme is optimized with parameters evaluation on statistical. The simulation results show that the proposed scheme can stabilize the blocking ratio within 15% in a normal traffic load scenario, and the relative optimization of blocking ratio is up to 27% better than that of the benchmark algorithm with time-sensitive requests.
What problem does this paper attempt to address?