Abstract

A promising way to realise a high capacity optical network is to utilise the wavelength dimension for address assignment of the nodes. In order to obtain scalability in such networks the wavelengths have to be re-used [1], [2] to allow the number of nodes to be independent of the number of available wavelengths. The idea is to divide the network into segments with equal wavelength maps and to connect these segments via wavelength converters.

© 1994 Optical Society of America