Power Control for Non-Gaussian Interference

This paper investigates a wireless communication system where the mutual user interference is not assumed to be a Gaussian process. We derive an exact expression for the average bit error probability (BEP) for such a system and study the non-Gaussian interference model through two types of power control problems. We analyze the situation under which the system can be asymptotically error-free, the behavior of users' BEP when scaling up a fixed power setting by a uniform scalar and the effect of varying symbol rate on the system performance. Our work shows that the non-Gaussian model has significantly different performance characteristics from the traditional Gaussian interference model. Simulations also show that the Gaussian model is generally pessimistic in comparison with the non-Gaussian model.