Research On Transmission Techniques In Multiple-Antenna And Cooperative Relay Networks

To improve the data rate and enhance the transmission reliability for the next-generation wireless communication systems, multiple-antenna and cooperative communications techniques have drawn much attention. On the one hand, multiple-input and multiple-output (MIMO) systems equipped with multiple antennas both at the transmitter and receiver can obtain the spatial multiplexing and diversity to improve the data rate and transmission reliability when combined with space-time codes. On the other hand, for some systems, in which it is difficult to equip multiple antennas considering the size, implementation complexity and power constraint of mobile terminals, single-antenna terminals located at different sites can also obtain spatial diversity by cooperative transmission. Cooperative communications can also improve the achievable rates of users, enhance the robustness of communication systems and save on resources for the whole network by cooperative terminals sharing resources. This thesis mainly addresses effective transmission techniques exploiting the spatial diversity in multiple-antenna and cooperative relay networks. Specifically, we mainly investigate how to compensate the diversity loss of generalized layered space-time codes due to the group detection in multiple-antenna systems and design distributed generalized low-density (GLD) codes for cooperative relay networks. The main contributions of the thesis are summarized as follows.Firstly, iterative group detection algorithms are proposed to reduce the diversity loss of generalized layered space-time codes due to the group detection in MIMO systems. Generalized layered space-time codes having the merits of layered space-time codes and space-time trellis codes, can obtain the higher data rate and enhanced transmission reliability. However, the group interference suppression and interference cancellation (IC) algorithm at the receiver also causes the diversity loss of some antenna layers. Minimum mean squared error (MMSE) and MMSE IC iterative group detection algorithms based on hard decision are proposed to avoid the diversity loss, reduce the error propagation and improve the system performance by utilizing the iterative interference cancellation. Moreover, MMSE IC iterative group detection also reduces the system sensitivity to the detection order. In addition, MMSE IC iterative group detection combined with power allocation among different antenna layers can further improve the system performance.Secondly, distributed generalized low-density codes are proposed for the clustered cooperative relay networks. In this scheme, GLD codes are decoded and forwarded in a distributed manner using the partial error-detecting and error-correcting capabilities of constituent codes in GLD codes. Distributed GLD codes can adapt well to the variation of the relay number due to the random mobility of nodes by adjusting the amount of constituent codes allocated to each relay and guarantee a fixed overall code rate if no relay becomes unreliable due to the decoding failure. For relay nodes, a low-complexity progressive processing algorithm is suggested to adapt to the source-relay channel conditions. At the destination node, the soft information from different paths is combined and sent to the GLD decoder for the iterative decoding, and thus diversity gain and coding gain are achieved simultaneously with low complexity. Distributed GLD codes can effectively combat quasi-static fading and achieve about tens of dB gain over the direct transmission. Furthermore, compared with other cooperative transmission schemes such as distributed Turbo codes, distributed GLD codes can also have some advantages such as the improved bit error rate performance, the lower complexity and the constant overall code rate. In addition, for a specific network topology, the effect of power allocation among the source and relay nodes is investigated considering the large-scale path-loss.Finally, for the universal cooperative relay networks, the probabilities of relay nodes becoming unreliable due to the source-relay channel fading and the effects on distributed GLD codes resulting from unreliable relays are analyzed. The probabilities of different number of relay nodes becoming unreliable are analyzed and the most likely number of relays tending to become unreliable is obtained for a specific network with the fixed number of relays according to the theories of Bernoulli trials. The relationship between the probability distribution of relay nodes becoming unreliable and the source-relay channel conditions is illustrated by simulations. Furthermore, distributed GLD codes with imperfect relay nodes are investigated and the effects on the system performance and the overall code rate resulting from unreliable relays are also analyzed. Simulation results verify that the performance of distributed GLD codes with some unreliable relays does not degrade significantly and the average performance approaches that with ideal relay nodes as the quality of source-relay channel improves.