A lot of solutions have been elaborated for data dissemination in self organizing networks. Every one of these can provide adequate performance in the environment, which it was designed for. There are adaptive solutions but these can not overcome the limitations in different solutions. The parameters of self-organizing networks depending on the environment can be a whole lot different : there are networks with hundreds of messages to deliver per day but there are networks with thousands of messages per second. The properties of mobile devices, their number, their mobility model, the supported wireless standards, the extension devices (i.e. : GPS) all require a different design approach. In general one can state that there is no protocol or network paradigm, which can provide the scalable, reliable and efficient functioning of self-organizing networks in every environment. The application of the evolutionary principle in different fields such as soft computing has brought us wonderful results. In my thesis I am investigating if it is possible to describe well-known protocols with good properties with a genetic programming language and then look at a group of protocols as a population, which consists of competing individuals. This way they can create new and better performing generations through reproducing and carrying the good properties of their ancestors with them. A system based on this idea can adapt to the most extreme conditions and can form itself everytime and everywhere on the given user requirements. It was an important goal in the
design process of the language to be easily extendable. With the help of the compiler we can easily test our new ideas in an evolutionary environment.