Coordinating Multiple Light-Trails in Multicast Elastic Optical Networks With Adaptive Modulation

• Published in: IEEE photonics journal Vol. 15; no. 1; pp. 1 - 15
• Main Authors: Cai, Anliang, Li, Yongcheng, Chen, Jian, Shen, Jianhua
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.02.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accommodation; Adaptive modulation; Adaptive systems; Algorithms; Benchmark testing; elastic optical network; Heuristic algorithms; Heuristic methods; Integer programming; Light; light-trail; Linear programming; Mixed integer; Modulation; multicast; Multicast algorithms; Multicast communication; Multicasting; Optical communication; Optical fiber networks; Routing; routing and spectrum assignment
• ISSN: 1943-0655; 1943-0647
• DOI: 10.1109/JPHOT.2022.3233304

Optical multicasting has been considered resource efficient for multicast services. Light-tree and light-trail are two technologies that support optical multicasting while the former requires many splitters and thus experiences significant power loss. In this paper, we consider using the light-trail technology for the accommodation of multicast requests in elastic optical networks with adaptive modulation. For better spectrum efficiency, we consider accommodating each multicast by multiple light-trails. We formulate the problem by Mixed Integer Linear Programming (MILP) and propose efficient heuristic algorithms. For the impact of accommodation sequence on the algorithm performance, apart from the traditional sequence among different requests, we consider an additional sequence among the destinations of a multicast. For efficient multicast accommodation, we propose several strategies and compare their performances through a range of cases. To avoid a destination occupying excessive resources in certain cases of joining multiple light-trails, we propose an efficient algorithm to delete some duplicated destinations. Numerical results show that the proposed heuristic algorithms significantly outperform a benchmark algorithm and one performs close to the optimal MILP. Also, the algorithm for deleting certain destination replicas largely reduces the spectrum and transmitter usages, up to 41% and 20% for the cases considered, respectively.