Jorge Cabrejas Peñuelas

Mobile communication systems are currently being developed with the aim of providing peak data rates up to 20 times higher to those of LTE-Advanced Rel 10. However, this performance improvement is often far from being the experimented performance by those users who are far from the Base Station (BS). In this sense, there exists a consensus on the fact that the best way to achieve the same quality for all users is with the use of heterogeneous networks composed of macrocells, microcells, femtocells, and relays. This dissertation addresses the use of Mobile Relays (MRs) to provide service to users who are at the cell-edge. MR is a natural extension of the fi xed relay in which users who are in the idle state could retransmit signals received from other transmitters to enhance data rates. This dissertation focuses on proposing and evaluating new techniques that manage the use of the MR in the new generation cellular networks. In particular, the dissertation studies MR from two complementary points of view. The first point of view investigates the MR management at the network level through a signaling protocol known as Media Independent Handover. The central idea of this mechanism is to use this signaling to connect the BS and the user in one of the following two manners. In the former, both entities are connected directly through the xG (x= 2, 3, 4, 5) wireless network. In the latter, there exists an xG connection between the BS and the MR and another one between the MR and the user through an IEEE 802.11 local wireless network. The investigations in this Thesis aim at fi nding a trade-of f between using multiple MRs and reducing signaling overhead. The second point of view deals with MR integration at air interface level. It consists in detecting, proposing, and evaluating new transmission techniques that solve the drawbacks derived from coherent detection. As with point-to-point systems, employing multiple antennas in a cooperative system can signi cantly improve the spectral efficiency of the systems with only one transmit antenna assuming that the channel estate information is available at the receiver. However, performing a coherent detection in a network assisted by relays consumes much more resources than a point-to-point network since the coherent detection requires the channel estimation of source-relay, relay-destination, and source-destination links. In this Thesis, the proposed solution is to use transmission techniques that do not need the channel knowledge to perform the detection. This dissertation evaluates the use of Single-User (SU) open-loop communication methods over temporally-correlated Rayleigh fading MIMO channels. On the other hand, in multi-carrier systems, the Thesis proposes to transmit the Grassmannian signaling (GS) in the virtual block formed by the coherence time and the coherence bandwidth. This proposal is due to the fact that GS achieves data rates approaching capacity over block-fading channels. However, this channel type is not common in real systems since channel correlation is often found in frequency, time, and space. For this reason, the next objective is to evaluate the performance of GS compared to the diversity transmission modes of LTE, analyzing the impact of user mobility and antenna correlation. Thanks to these investigations, we point that non-coherent systems are promising techniques in mobility scenarios with a high number of transmit antennas. This result motivates its relevance in the design of new SU open-loop transmission methods with multiple antennas. In downlink multi-user non-coherent scenarios, superposition coding and a suboptimum detection scheme are proposed. This detection system reduces the complexity respect to the maximum likelihood detection. Finally, this dissertation proposes that GS is transmitted in a new carrier type, where any reference signal is transmitted. In this way, the user would change its detection method to non-coherent.