In this thesis we investigate the effects of fixed and fixed alternate routing on the wavelength conversion (WC) gain in wavelength-routed all-optical networks. Hence, we compare the benefits of (WC) with those obtained from increasing the number of alternate routes (K). Furthermore, we propose a novel heuristic algorithm, Highest Node Blocking (HNB), to solve the NP-complete wavelength converter placement problem. We find that providing more (K) in a mesh topology improves the performance better than wavelength converters do. However, the converters' gain emerges as K increases to a point when placing a converter improves the performance and increasing K worsens it. The results, based on extensive simulations, show that it is important to determine the advantage turnabout point between increasing K and placing wavelength converters. Finally, for most simulated networks and traffic patterns, we find that our heuristic outperforms previously proposed algorithms. |