Statistical inference for high dimensional parameters in communication systems

Assessment of performance of communications systems generally translates to the statistical problem of estimating a high dimensional parameter. Inference about such parameters is often complicated by the fact that data from communication systems exhibit complex dependencies and are rarely easily cast in a simple parametric model. The aim of the PhD project is to develop mathematically rigorous methods to approach these issues in the context of real-world problems. Resampling methods, asymptotic statistical inference and statistical inference for dependent data are keys in this connection. A number of projects have been formulated and are now studied e.g. 1) resampling methods for the class of regenerative stochastic processes with applications to nonparametric inference for queuing systems, 2) development of asymptotic results for location estimation in wireless multihop networks and 3) identification of nonparametric inverse problems in general models for queuing systems.