Over the last decade, translucent WDM networks have appeared as a promising candidate for next generation core networks. Using sparse regeneration techniques, translucent networks may achieve a pretty tradeoff between the low cost of transparent networks and the quality of transmission guaranteed by fully-opaque networks. On one hand, deploying large-scale transparent networks is still a critical issue since transmission impairments arising from long-haul optical equipment may significantly limit the optical reach. On the other hand, opaque networks remain very expensive due to electrical regeneration performed at each network node. In this paper, we propose an original exact approach, based on an integer linear program (ILP) formulation, to deal with the problem of translucent network design. Existing exact approaches rely on linear approximations of the signal degradation. In this paper, we make use of a realistic estimate of the signal quality taking into account the simultaneous effect of four well-known transmission impairments. Moreover and to the best of our knowledge, all existing approaches consider the problem of translucent network design assuming either permanent or semi-permanent lightpath demands. In this paper, we consider the problem of translucent network design under dynamic but deterministic traffic pattern; i.e., scheduled lightpath demands (SLDs). In order to improve the scalability of our approach, we decompose the problem into the routing and regenerator placement, and the wavelength assignment and regenerator placement sub-problems. In the former, we place regenerators and route demands while assuming that the quality of transmission is independent of the wavelength value. In the latter, additional regenerators may be required to overcome the dependency of the quality of transmission on the wavelength value. Deployed regenerators may be shared among multiple non-concurrent SLDs. In doing so, we shorten further the gap between translucent and - - transparent network costs.