Comunicaciones móviles de misión crítica sobre redes lte

  1. García Pérez, César Augusto
Supervised by:
  1. Pedro Merino Gómez Director

Defence university: Universidad de Málaga

Fecha de defensa: 20 July 2020

Committee:
  1. Carlos García Rubio Chair
  2. Mercedes Amor Pinilla Secretary
  3. Fidel Liberal Malaina Committee member

Type: Thesis

Teseo: 629295 DIALNET lock_openRIUMA editor

Abstract

Mission Critical Communications (MCC) have been typically provided by proprietary radio technologies, but, in the last years, the interest to use commercial-off-the-shelf mobile technologies has increased. In this thesis, we explore the use of LTE to support MCC. We analyse the feasibility of LTE networks employing an experimental platform, PerformNetworks. To do so, we extend the testbed to increase the number of possible scenarios and the tooling available. After exploring the Key Performance Indicators (KPIs) of LTE, we propose different architectures to support the performance and functional requirements demanded by MCC. We have identified latency as one of the KPI to improve, so we have done several proposals to reduce it. These proposals follow the Mobile Edge Computing (MEC) paradigm, locating the services in what we called the fog, close to the base station to avoid the backhaul and transport networks. Our first proposal is the Fog Gateway, which is a MEC solution fully compatible with standard LTE networks that analyses the traffic coming from the base station to decide whether it has to be routed to the fog of processed normally by the SGW. Our second proposal is its natural evolution, the GTP Gateway that requires modifications on the base station. With this proposal, the base station will only transport over GTP the traffic not going to the fog. Both proposals have been validated by providing emulated scenarios, and, in the case of the Fog Gateway, also with the implementation of different prototypes, proving its compatibility with standard LTE network and its performance. The gateways can reduce drastically the end-to-end latency, as they avoid the time consumed by the backhaul and transport networks, with a very low trade-off.