Shannon's Channel Capacity has long been an elusive and merely theoretical goal, with real code structures and hardware implementations providing performance relatively far from the limit. With the somewhat recent advent of near-capacity-achieving codes, however, we can now actually use capacity calculations as practical metrics in optical link designs. In link budget calculations, we can explicitly show losses with respect to capacity that are directly traceable to engineering choices such as sub-optimum code rate selection, sub-optimum code structures, sub-optimum decoding architectures, and other effects. In this way, engineering elements of modulation and coding design can be compared equally with compromises in optics, tracking, and so on. We can further use capacity calculations to predict performance in fading channels, the bane of atmospheric and imperfectly tracked optical links. Such analysis suggests structures using coding and possibly interleaving that can get very close to the optimum performance. In fact, performance should be related to the average fade depth and not the deepest fades.
展开▼
