Departmental lecture: Asymptotics of large communication systems

Benjamin Meaker visiting Professor: Peter Smith

Date: 26 July 2018
Time: 2pm
Location: Howard House, Level 4 Seminar Room

Large wireless communication networks are widely believed to offer
“free” interference reduction. This has led to a huge growth of research
in performance analysis of large systems, where many users communicate
with 100s of antennas located in a large array of antennas. Since
wireless channels are often modelled as random Gaussian matrices, a
considerable body of work now exists on the effect of large random
matrices in performance analysis. The process where large numbers of
antennas automatically removes interference is called favourable
propagation and is a property of large random vectors. However, in
contrast to the statistical Gaussian channel models, more fundamental
channel models are also of interest based on physical ray-tracing. In
this talk I demonstrate that the two types of models lead to very
different conclusions. In order to do this, the asymptotic interference
variable is analysed as the numbers of antennas and users in the system
grows to infinity.