Statistical Physics in Advanced Multiuser Communications

Statistical physics and multi-user communications show strong analogies from a conceptual point of view. In both cases many objects interact with each other through variables that are constrained to be binary. These interdisciplinary analogies can be expl oited to advance the understanding and design of future wireless communication systems. Though the analogies between the two fields do not extend too far and, in real-world communication systems, statistical physics results cannot be applied directly, the engineering community can strongly benefit very much from the analytical toolboxes developed by physicists to understand the macroscopic properties of spin glasses. The research focuses on improving the efficiency of wireless multi-user communications, t hus reducing the demand of physical spectrum, prolonging the stand time of battery driven devices, and reducing the emitted electromagnetic radiation. To achieve this goal the project will target several important open problems in wireless multi-user comm unications: • Capacity analysis of multi-user systems with dual antenna arrays in correlated fading. • Speed of convergence of large system results for non-linear multi-user detectors towards their asymptotic limits. • Impact of detector mismatch on the p erformance of wireless multi-user receivers. • Pre-coding for wireless broadcast channels. Additionally, the project seeks to achieve a new realm in education for communication engineers: The project will develop interdisciplinary course programs to incre ase mutual exchange between physicists and communication engineers. The statistical mechanics interpretation of digital communications will be reformulated into a form that is suitable for students with a standard engineering background to give them new i nsight and additional intuition when designing communication systems. An experimental graduate program will be implemented at NTNU.