Stochastic multipath propagation models for indoor and outdoor mobile radio channels are in general derived by employing coloured Gaussian noise processes. Efficient design and realization techniques of such processes are therefore of particular importance in the area of mobile radio channel modelling. This paper presents especially for complex coloured Gaussian noise processes a deterministic simulation model that is based on Rice's (1944) sum of sinusoids. A new method is introduced for the derivation of an optimal set of sinusoids, i.e., the procedure is such that for a given number of sinusoids an optimal approximation of Gaussian (Rayleigh, Rice) processes with given Doppler power spectral density shapes is obtained. Moreover, the statistical properties of the resulting (deterministic) simulation model are investigated.