Leveraging Context Information in Fifth Generation MillimeterWave Mobile Networks: a Bayesian Approach

In order to meet their ambitious data-rate targets, future 5th generation mobile communication systems will enable communication over the millimeter-wave frequency bands (30-300 gigahertz). At these frequencies, the use of highly directive antennas is needed in order to overcome the large attenuation that the signals suffer over distance. This will be accomplished by using steerable antenna arrays combined with sophisticated signal processing methods that steer the involved antenna arrays in the required directions: these are typically called beamforming methods.

In this project, we aim to use Bayesian signal processing methods to design novel beamforming algorithms that exploit context information available in the system: locations of transceivers, their direction of movement and orientation, and information on the propagation environment. In this way, we aim to obtain algorithms that have better performance and are more robust than their context-agnostic counterparts.