| By retransmitting information for one another, single-antenna terminals can form a virtual antenna array such that the end-to-end transmission reliability can be significantly strengthened. Such an approach to achieve spatial diversity by means of sharing transmit antennas is called cooperative diversity. From the very beginning, cooperative diversity techniques have been facing two fundamental problems: one concerns how to attain full diversity and to improve spectral efficiency simultaneously, and the other one is how to achieve full diversity at the expense of a lower signaling overhead and a shorter feedback delay. On the other hand, for traditional cooperative systems, the role for each terminal(i.e., as a source or relay) is generally fixed, which is harmful to a flexible system configuration and to an equal opportunity to transmit information. In view of these issues, in this dissertation we make an in-depth investigation into the foregoing problems and our main contributions are outlined as below: 1. The study of efficient diversity exploitation methodologies for multiuser cooperative systemsTwo types of high-efficient diversity exploitation measures are proposed for multiuser cooperative systems, which includes two-step exploitation measures and incremental exploitation measures. Afterwards, these measures are respectively applied into multi-source, multi-relay cooperative systems and single-source, single-relay, multi-destination cooperative systems. More specifically,(1) For multi-source multi-relay cooperative systems, a two-step measure to exploit multiuser diversity and cooperative diversity is proposed. Under AF(Amplify-and-Forward) and DF(Decode-and-Forward) relaying modes, the two-step(source-relay) selection schemes are respectively presented to achieve full diversity, which also alleviates the difficulty to exploit diversity gains in multi-source, multi-relay cooperative systems. Both theoretical analysis and numerical results manifest that: two-step selection schemes can still attain full diversity and the required amount of channel state information(CSI) decreases from MN+M+N to M+2N, where M and N respectively denote the number of sources and relays in the systems. Also, the distributed timer techniques can be used to perform the node selection, which considerably lowers the implementation complexity.(2) For single-source, single-relay, multi-destination cooperative systems, the Incremental DF relaying scheme with MUltiuser diversity(MU-IDF), which makes good tradeoff between transmission robustness and spectral efficiency, is proposed. By adopting an idea of second-round opportunistic scheduling, we present an improved spectrally-efficient diversity exploitation scheme with higher transmission robustness, namely, MU-IIDF(Improved Incremental DF relaying with MUltiuser diversity). Differently from traditional cooperative schemes, the proposed schemes can recover full diversity gain and significantly improve the spectral efficiency. In the high SNR regime, the spectral efficiency of the proposed schemes approaches that of direct transmission.(3) In view of the fact that the ANC(Analog Network Coding) protocol cannot achieve full diversity and the TDBC(Time Division Broadcast) protocol suffers from low spectral efficiency, we combine multiuser diversity and incremental relaying to design a more efficient two-way relaying protocol, which attains a superior diversity-multiplexing tradeoff(DMT) than ANC and TDBC. In particular, the outage performance and achievable DMT of the proposed scheme is strictly analyzed. Compared with TDBC, the proposed scheme achieves a higher spectral efficiency and harvests a higher diversity order than ANC. The significance of this work lies in that: the DMT performance of the proposed two-way relaying scheme is so far the best one. 2. The study of distributed link/antenna selection schemes for cooperative systemsIn multi-antenna cooperative systems, the exploitation of spatial diversity often involves the opportunistic selection of link and transmit antennas. Traditional schemes usually adopt a centralized approach to perform the link/antenna selection, which requires the centralized collection of global CSI, leading thus to considerable signaling overhead and selection delay. To address these issues, we formulate a local CSI comparison and decision feedback mechanism(namely, the distributed decision feedback mechanism), and then proposes the distributed link/antenna selection schemes. Compared with the centralized selection schemes, the most favorable advantage of the distributed selection schemes is that: besides lowering the signaling overhead and feedback delay, the proposed distributed selection schemes can achieve almost the same transmission reliability with that of the centralized ones. In this dissertation, the advantages of the proposed distributed decision feedback mechanism are validated through two kinds of classical multi-antenna cooperative systems.(1) Based on the proposed distributed decision feedback mechanism, this part designs distributed link selection rules and the corresponding flowcharts for variable-gain and fixed-gain AF relaying systems with transmit beamforming. In addition, the transmission robustness and implementation complexity of the proposed schemes are compared with the counterparts of the centralized ones. Numerical results manifest that the proposed distributed link selection schemes can achieve almost the same outage performance with that of the centralized schemes over the entire SNR regions. Moreover, when the relay is deployed around the mobile station, the proposed scheme can be implemented in a nearly perfect distributed fashion in addition to possessing a high transmission robustness.(2) Based on the proposed distributed decision feedback mechanism, this part presents an idea of distributed antenna selection and then designs the distributed antenna selection rule and the key implementation steps. Compared with the outage-optimal centralized antenna selection measure, DAS can still achieve full diversity and maintain a much lower implementation complexity. Moreover, unlike the existing antenna selection schemes, the signaling overhead of DAS does not increase with the number of transmit antennas. In addition, numerical results indicate that the transmission reliability of DAS is very close to that of the optimal centralized antenna selection scheme when the relay is deployed in an outage-optimal manner. 3. The study of opportunistic role selection mechanism and its implementation issuesIn view of the drawbacks that in the existing cooperative systems, the role for each terminal(i.e., as a source or relay) is generally fixed, this part proposes an idea of ROle SElection(ROSE), and then develop a Centralized ROSE(C-ROSE) scheme, a Distributed ROSE(D-ROSE) scheme, and a Proportional fair scheduling based ROSE(P-ROSE) scheme for two-user AF relaying systems. The three ROSE schemes can respectively achieve the optimal system outage performance, the lowest signaling overhead, and the fairest role selection, which significantly improves the fairness and transmission robustness between the two users. Both theoretical analysis and simulation results show that: a) the high-SNR outage performance of the three ROSE schemes is exactly the same and is dominated by the two user-to-destination links irrespective of the inter-user link; b) under the scenario of balanced user-to-destination links, the outage performances of the three ROSE schemes are nearly the same over the entire SNR regions; c) under the scenario of unbalanced user-to-destination links, the outage performance of P-ROSE is inferior to that of C-ROSE and D-ROSE especially for the case of a weak inter-user link.
|
PDF Full Text Request