PhD defense of Juan Carlos Vargas Rubio on March 23th, 2023

The defense of PhD thesis of Juan Carlos Vargas Rubio will be held on March 23th, 2023 at 10:00 in the Rennes campus of IMT Atlantique (Petit amphi). It will be accessible with a conf call. The thesis is entitled “Unicast versus Broadcast in Cellular Networksn 5G Radio Access Networks”. It was made in the framework of a CIFRE contract within Enensys.

Data traffic on mobile networks increases every year, especially video content. However, spectrum is scarce and expensive and operators need to optimize its use. In scenarios where the same content is transmitted at the same time to many devices in the same geographical area, the preferred solution to reduce bandwidth consumption is broadcast transmission.
Unicast transmission benefits from link adaptation techniques. However, the same content is transmitted as many times as the number of users demanding the same service. Conversely, a single broadcast transmission can cover a large number of users. Nevertheless, the bitrate in broadcast is fixed considering the users with the worst channel quality. Multicast-Broadcast Single-Frequency-Network (MBSFN) is a broadcast technique in which a group of synchronized cells transmit the same waveform. On the other hand, with Single-Cell Point-To-Multipoint (SC-PTM) each cell performs broadcast transmission independently. The problem is to determine when is it better to use unicast, MBSFN or SC-PTM.
In our work, we compare the performance of unicast, MBSFN and SC-PTM through system level simulations and analytical models. We consider base stations located according to Poisson distributions, the use of beamforming in unicast and different broadcast configurations. Furthermore, we propose an analytical method to calculate the number of users demanding the same content from which MBSFN or SC-PTM become more efficient than unicast. We prove that a switching mechanism based on this user threshold reduces bandwidth utilization and energy consumption. This method is based on stochastic geometry results for wireless networks.