Abstract
The routing and spectrum assignment (RSA) problem has emerged as the key design and control problem in elastic optical networks. Distance adaptive spectrum allocation exploits the tradeoff between spectrum width and reach to improve resource utilization by tailoring the modulation format to the level of impairments along the path. In this paper, we consider the distance-adaptive RSA (DA-RSA) problem with fixed alternate routing. We first show that the DA-RSA problem in networks of general topology is a special case of a well-studied multiprocessor scheduling problem. We then leverage insights from scheduling theory to (1) present new results regarding the complexity of the DA-RSA problem, and (2) build upon list scheduling concepts to develop a computationally efficient solution approach that is effective in utilizing the available spectrum resources.