During the last two decades, wavelength conversion has received a considerable attention due to its strong influence on the blocking performance of wavelength-routed WDM networks. However, most of the related works focus on only one aspect of wavelength conversion, i.e., sparse or partial or limited wavelength conversion. Although these approaches outperform the full wavelength conversion, there has been no fair comparison among different approaches. Furthermore, it is still arguable which wavelength conversion scheme leads to the optimal performance of such networks. Recognizing these limitations, we advocate the use of sparse-partial-limited wavelength conversion (SPLWC) integrating all above approaches. In this paper, we consider a converter placement problem of minimizing the wavelength conversion cost (WCC) to meet the constraint on the blocking probability. We present a novel analytical model accounting for the two sources of call blocking in wavelength conversion: a range blocking from the limited conversion range of a wavelength converter; and a capacity blocking from the limited number of wavelength converters. We also present a sequential converter placement algorithm which can be applicable to any wavelength conversion schemes including SPLWC. From the numerical results, we demonstrate that the blocking performance of the analytical model closely matches with that of the simulation. We also show that SPLWC achieves an outstanding WCC performance compared with the existing approaches.
展开▼
