Comparative Ability of Span-restorable and Path-protected Network Designs to Withstand Error in the Demand Forecast

– The growth of the Internet, wireless mobility, and economic variability all contribute to the difficulty of forecasting the demand for which an optical transport network should be planned. It is therefore of interest to know how various architectural options inherently withstand changes in demand pattern relative to a forecast. In this study, we focus on the span restoration (SR) and shared backup path protection (SBPP) schemes in the sense of how well they withstand departures of the actual demand pattern from the forecast demand to which they were designed. As a measure of the severity of error in a given forecast, we propose a metric called Pattern Forecast Accuracy (PFA) which seems to more directly reflect the number and magnitude of demand errors than prior correlation measures. We also define the maximum servability optimization problems for SR and SBPP to assess their fundamental robustness to forecast errors. Based on results with three test networks under several thousand mis-forecast demand trials, we find that the minimum-capacity SR designs produced to the nominal forecast usually serve (and protect) more of the mis-forecast demands than the corresponding SBPP designs. But the largest differences are under 10%, and average only about 3%. The preliminary conclusion is therefore that the schemes are essentially equally “future-proof.” Further work should, however, consider how jointlyoptimized working path routing and small margins of extra capacity may improve the robustness to forecast error. The basic framework introduced here, of PFA to characterize forecast error, and maximum servability sub-problems to assess robustness, is suggested as a general methodology for quantitative studies of the intrinsic robustness of various transport architectures.