The project aims at the development of efficient mathematical techniques for the analysis of stochastic models that are useful for the performance analysis of modern heterogeneous telecommunciation networks. These models stem from queueing theory and estimation and decision theory. They are designed to provide more realistic characterizations for information sources and arrival streams, service mechanisms and scheduling techniques in network nodes, as well as for the study of non-classical aspects of performance evaluation such as the incorporation of hardware delays and information deadlines, and the analysis of loss processes and transient behavior. The models will also be applied in the context of adaptive modulation and coding on feedback channels, iterative carrier and clock synchronization, signal design for localization systems, etc.