Lightpath Rerouting Strategies in WDM All-Optical Networks Under Scheduled and Random Traffic

In WDM all-optical networks without wavelength conversion, traffic rerouting is motivated either by an optimization of resource utilization or by network survivability. In this paper, we use rerouting to optimize network resources allocation in order to set up an incoming lightpath demand to be blocked for lack of resources. Rerouting aims at reassigning the wavelength and/or the path of one or several established connections in order to free enough wavelengths to satisfy the incoming demand. Rerouting refers implicitly to dynamic traffic. In most previous studies related to rerouting, only random (dynamic) traffic is considered. In this paper, we propose a new lightpath rerouting scheme considering three types of traffic demands, referred to as permanent lightpath demands (PLDs), scheduled lightpath demands (SLDs), and random lightpath demands (RLDs). PLDs are static, whereas SLDs and RLDs are dynamic. SLDs are preplanned, whereas RLDs are stochastic. PLDs may be seen as a particular case of SLDs. PLDs and SLDs correspond to guaranteed services, whereas RLDs correspond to best-effort services. Thus, PLDs and SLDs cannot be rerouted. We here describe two new routing and wavelength assignment (RWA) strategies applying rerouting. Both strategies assume that PLDs are routed offline during the network planning phase. The first strategy computes the RWA for SLDs and RLDs on the fly. The second strategy proceeds in two separate phases. It first computes offline the RWA for SLDs before considering RLDs on the fly on the remaining network resources. Our rerouting schemes aim at minimizing the number of RLDs or the number of optical channels to be rerouted. Through numerical examples and experimental simulations, we outline that routing SLDs offline and RLDs online instead of routing SLDs and RLDs online enables lower rejection ratios. We also compare our proposed rerouting algorithms with other approaches from the literature in terms of complexity.