Transparent Path Length Optimized Optical Monitor Placement in Transparent Mesh Networks

Abstract

Linear programming techniques are applied to the optimization of optical monitor placement by minimizing the worst case transparent path length in optical networks, which we study over different network topologies and traffic patterns.

Impact of wavelength route correlation on the optimal placement of optical monitors in transparent mesh networks

“…The route-correlated curves are systematically below their uncorrelated counterparts, indicating a significant reduction in monitoring cost. Furthermore, the correlated cases deviate from the inverse relationship (linear on log-log plot) observed previously for the uncorrelated cases [2]. Different loading parameters bore out the same pattern, but we also observed a variation in the separation between the two sets of curves.…”

“…As previously [2], the repair cost is given by the effective transparent path length, which is the number of un-monitored links around a fault location along the impacted lightpaths -representing the fault location uncertainty due to transparency. We optimise to avoid the worst case or maximum repair cost.…”

“…A range of uniform demand, regular networks with variable mean degree and traffic loading [2] were used to compare uncorrelated and correlated procedures. Shown in Figure 2 are results for a medium-loaded mesh network (30 biconnected addnode nodes, with 3 and 6 optical amplifiers per A/D hop), showing a maximum repair cost (size of unmonitored transparent region) plotted on a log-log graph, against monitoring cost (number of monitors placed), with hop lengths per wavelength varying across the set {2,4,6,8}, averaging over 10 random loading instances.…”

“…For this application, OPM is used to selectively monitor the optical signal quality along the propagation path in order to effectively reduce the size of transparent regions. Recently we studied the optimisation of monitor placement with respect to reducing the transparent path length between monitoring sites [2]. This approach starts with the assumption of perfect monitoring and explores the impact of topology and traffic on the monitor placement strategy.…”

Optimizing the monitoring path design for independent dual failures

“…Specifically, optical performance monitoring is the basis for fast failure detection and localization, which are crucial steps in the fault management process [2]. Accurate identification of failed elements mitigates the range and duration of failure impacts, reduces the operational cost induced by on-site physical repairs, and justifies the capital cost to deploy the optical performance monitors [3]. In particular, the faulty-element information is required to initiate the connection recovery processes for any failure-dependent/failure-specific recovery schemes, such as pcycle, link protection, restoration, etc.…”