Reverse manycast data retrieval in Elastic Optical Networks

Extreme-scale science applications are highly innovative and constantly evolving. They are expected to generate data in the petabyte and exabyte ranges. This data needs to be transferred, processed, and analyzed at remote locations. A flexible data retrieval service is needed, where a user requesting data retrieval from a remote site can have a choice between replicated storage sites. Elastic Optical Networks are ideal backbone networks, since they can efficiently utilize the optical fiber's bandwidth in an elastic manner by partitioning the bandwidth into hundreds or even thousands of subcarriers. In this paper, multi-sourced data retrieval problem, called reverse manycast, is studied for static traffic in elastic optical networks, the objective is to minimize the total transmission completion time of all the requests. A novel ILP formulation and a low-complexity heuristic are proposed. Simulation results are presented to demonstrate that the proposed method can save up to 37% in completion time compared with a benchmark.