Research On Superposition Coding Based Non-orthogonal Transmission

Interference will deteriorate the performance of wireless communications. The current wireless communication systems adopt orthogonal signaling to avoid the interference among the users, which though simplifies the system design, but results in inefficient use of the channel resources. From an information-theoretic viewpoint, non-orthogonal transmission instead of the orthogonal one can maximize the efficiency of wireless communications. Superposition coding based non-orthogonal transmission has achieved tremendeous attention in recent years. Superposition coding in conjunction with successive interference cancellation(SIC) allows multiple users(or date flows) to share the same channel resource at the same time, for which there is no need for the user signals to be orthogonal with each other, and hence more efficient channel access can be realized. In this dissertation, superposition coding based non-orthogonal transmission is studied with the following three aspects being taken into account: the combination of superposition coding and relaying technology, superposition coding related optimal designs, and system-level optimization. It is indicated that in superposition coding based non-orthogonal wireless communication systems, the different users(or data flows) form an alliance and can accomplish their communication task collaboratively, which achieves higher transmission efficiency than orthogonal wireless communications. The main contributions and innovations of this dissertation are summarized as follows:1)Superposition coding based relaying protocol design for single-user relay systems. A novel superposition coding based partial forward(SCPF) relaying protocol and a time compress and forward(TCF) relaying protocol are proposed for the case with a relatively better source to relay(S-R) link and the case with a relatively better relay to destination(R-D) link, respectively. The disparity in the qualities of the links consisting of a relay channel has a direct effect on the efficiency of a relaying protocol. By using superposition coding based non-orthogonal transmission, we can design protocols that adapt to such link disparity to avoid transmission efficiency loss due to the mismatching between the traffic and the quality of a link. Given that repetition coding is adopted by the relay, the proposed protocols can neatly adjust the traffic of the links. In the SCPF relaying protocol, the source transmits an additional message to be directly decoded by the destination at the same time when it is transmitting the message to be forwarded by the relay, which benefits from the relatively better quality of the S-R link. In the TCF protocol, multiple received signals at the relay are superimposed first and then forwarded simultaneously, which can benefit from the relatively better quality of the R-D link. Methods for performance analysis of the SCPF and the TCF relaying protocols are provided. Theoretical analyses and simulation results indicate that in the corresponding scenarios, the proposed protocols can evidently improve the efficiency of relaying.2)Performance analysis and design of superposition coding based non-orthogonal multiple access(NOMA). For the case with only statistical channel state information(CSI), we proposed a method to balance the outage probabilities of the users, which privides an analytical method for the outage performance evaluation of downlink NOMA systems when user fairness is considered. First, we investigate the selection of the power allocation factors(PAFs) and decoding order for superposition coding based non-orthogonal transmission, and derive the constraints that should be satisfied by the optimal PAFs under a certain assumed decoding order, the decoding order that achieves the optimal user fairness, and the closed-from PAFs and outage probability expressions of the users at the optimal solution of the outage balancing problem. Then, based on the above results, we investigate the problem of user grouping, including the grouping approach and the inter-group power and channel resource allocation. By analyzing the KKT optimality conditions, we propose an algorithm to solve the optimal power and channel resource allocation among different user groups. The implementation complexity issue of NOMA is also considered with focus on that caused by SIC and user grouping. We use the average of the balanced outage probability over a larger number of randomly generated scenarios as a performance measure, and analyze the impact of different user grouping approaches, the group size, and inter-group power and channel resource allocation on the performance of NOMA. It is shown that user grouping serves as an effective method to achieve a good tradeoff between the complexity and performance of NOMA, and with a moderate complexity increase, NOMA performs much better than conventional orthogonal multiple access.3)Analysis and optimization of a superposition coding based pairwise transmission strategy in downlink relay networks. When compared with downlink networks without relay, the spatial differences from the relay to different users form a new freedom that can be exploited by superposition coding based non-orthogonal transmission to achieve performance gain. Moveover, the bandwidth loss of relaying duo to half-duplex constraint can be compensated by non-orthogonal multiplexing. To make the transmission strategy practical to be applied, we consider a superposition coding based pairwise relaying protocol where the relay adopts repetition coding and works in decoded and forward mode. With the pairwise relaying protocol, the source can communicate to two users simultaneously with the help of the relay in the same time-frequency channel. First, for given transmission rates of the users, the optimal selection approach of the PAFs and the decoding order for the pairwise relaying protocol is formulated, and the analytical outage probability expression of each user is given. Then, by high signal to noise ratio(SNR) analysis, the power minimization problem under the quality of service(QoS) requirements of the users is solved, where the QoS contains the required transmission rate and outage probability of a user. The corresponding optimal PAFs and decoding order are also obtained. Further, we investigated how to pair the users to minimize the overall power consumption of the system, which is formulated as a non-bipartite optimum matching problem. An algorithm with computational complexity of O(K2) is provided, which achieves a performance that is very close to that of the optimal matching which can be solved in polynomial time O(K3), where K is the number of the users. The simulation results show that when compared with conventional relaying scheme, the pairwise relaying can provide considerable power gain, and the high SNR approximation coincides very well with the numerically obtained exact results.