For speech recognition in mobile environments a number of severe problems have to be solved. This paper gives an overview of several approaches dealing with the basic problem of noise handling in speech recognition in relation to the characteristica of noise in the mobile environment. Using a qualitative description of the noise characteristica, it can be seen that most of the currently investigated noise-handling strategies are just capable of handling parts of the phenomena of "mobile noise". From the analysis of these algorithms a general architecture of a noise resistent speech recognition strategy for further investigation is derived.