JMAWR: joint optimization of monitoring trail allocation and wavelength routing with limited monitoring resources

In all-optical networks, monitoring trail (m-trail) has long been proposed as an effective way for link failure localization. Previous works tried to minimize the number of used m-trails for localizing network-wide single link failures, and they all supposed that the monitoring resources were enough. However, failures are rare events and the traffic demand is increasing rapidly. The dedicated wavelengths for m-trails cannot be used to transmit data traffic, and some traffic demand may not be served during the peak period. Considering the operators prefer to serve customers’ demand as much as possible, if the resources allocated for m-trails are limited, more traffic will be carried. In this paper, we focus on the scenario where there are not enough resources for allocating m-trails. In this scenario, as only part of single link failures can be unambiguously localized, we aim to monitor the most valuable links, which carry more traffic demand, and minimize the maximum number of used wavelengths for traffic demand on the links which cannot be unambiguously localized. As we can also control the wavelength routing for traffic demand, we try to jointly optimize the m-trail allocation and wavelength routing. We first formulate this joint optimization problem as a mathematical model and present concrete analyses on this model. According to these analyses, a joint m-trail allocation and wavelength routing (JMAWR) heuristic algorithm is proposed based on graph-constrained group testing. Simulation results show that JMAWR outperforms the state-of-the-art m-trail allocation algorithm by up to 88.46% improvement in terms of maximum number of used wavelengths for traffic demand on links which cannot be unambiguously localized.